UNIVERSITY OF CINCINNATI
Date: 12-Nov-2009
I, Colin P White , hereby submit this original work as part of the requirements for the degree of: Master of Science in Biological Sciences It is entitled: Molecular Microbial Ecology and Operational Evaluation of a Full-scale and
Pilot-scale Biologically Active Filter for Drinking Water Treatment
Student Signature: Colin P White
This work and its defense approved by: Committee Chair: Ronald Debry, PhD Ronald Debry, PhD
6/21/2010 273 Molecular Microbial Ecology and Operational Evaluation of a Full- scale and Pilot-scale Biologically Active Rapid Sand Filter for Drinking Water Treatment
A thesis submitted to the Graduate School
of the University of Cincinnati
in partial fulfillment of the requirements for the degree of
Master of Science
In the Department of Biological Sciences of the McMicken College of Arts and Sciences
By
Colin Patrick White
B.S. Biological Sciences
University of Cincinnati April 2007
Committee Chair: Ronald W. DeBry, Ph.D. Abstract
Nitrification in drinking water distribution systems is a problem prevalent throughout the world, and it has become more pertinent since chloramination has become a popular disinfectant technique. Because nitrification requires ammonia, removing ammonia in source waters prior to treatment would benefit both the utility and consumers. Biologically active filtration is a well known technology in Europe but its reliability, and thus implementation, is questioned in the United States. In this study, natural microbial flora from a full-scale treatment plant in Greene County, Ohio was used to seed two pilot scale rapid sand filters. These filters were evaluated for their ability to oxidize ammonia-nitrogen. Molecular techniques, including 16S ribosomal RNA and amoA gene sequencing and denaturing gradient gel electrophoresis (DGGE) analysis, were used to phylogenetically identify and fingerprint the isolates. In addition to investigating nitrification, microbial arsenic oxidation was also investigated in pilot-scale filters.
Chemical analysis and microbial ecology is compared and discussed in terms of operational changes and water chemistry.
2 Dedication
To my parents
Acknowledgements
First, I would like to thank my advisory committee in the Department of
Biological Sciences: Drs. Ronald DeBry and Jodi Shann, for their support, guidance, and patience over the years.
Next, I would like to acknowledge the Environmental Protection Agency, Office of
Research and Development for which funding for this research via the UC/EPA
Traineeship was awarded. Special thanks goes to Dr. Darren Lytle, who arranged for this opportunity and was an invaluable resource as well as a great mentor. Under his expert tutelage I gained an appreciation for collaborative and applied research. Without his mentorship and constant encouragement I would never be where I am today. Credit is also due to members of the US EPA Treatment Technology Evaluation Branch, especially
Daniel Williams and Christy Muhlen, for their endless help and guidance both in the lab and in life.
In addition, I would like to thank my laboratory mates, Amy Reed, Andrea
Galloway, Alissa O’Donnell, Jermaine Conover, Jennifer Liggett, and David Hauber for their help and friendship.
3 TABLE OF CONTENTS
List of Tables
Table 3.2.1. Sample/Monitoring Strategy…………………………….………….34
Table 3.2.2. Sampling/Analytical Procedures…………………………..………..35
Table 3.5.1. Primers and References to Reaction Conditions……………………39
Table 3.5.2. PCR Reaction Setup………………………………………………..40
Table A1. QA Objectives for Method Detection Limit, Precision,
Accuracy, and Completeness…………………………………………………...108
Table A2. Summary of QC Checks…………………………………………….110
List of Figures
Figure 1. Nitrogen Fertilizer Loading in the United States……………………….15
Figure 2. USGS National Water Quality Assessment of Filtered
Ammonia in Source Water………………………………………………………..16
1. Introduction
1.1 Background…………………………………………………………………….6
1.2 Objectives……………………………………………………………………...9
1.3 Approach……………………………………………………………………...9
1.4 Overview of Thesis………………………………………………………..…10
2. Literature Review
2.1 Overview of Ammonia Behavior in Water…………………………………...11
4 2.1.1 Ammonia Chemistry……………………………………………….11
2.1.2 Ammonia Distribution……………………………………………..13
2.2 Ammonia Health Effects……………………………………………………..16
2.2.1 Carcinogenic Effects and Risks…………………………………….16
2.2.2 Non-Carcinogenic Effects and Risks……………………………….17
2.2.2.1 Acute Exposure………………………………………….17
2.2.2.2 Chronic Exposure………………………………………..17
2.2.3 Beneficial Effects…………………………………………………..18
2.3 Treatment Technology Options……………………………………………..18
2.3.1 Overview…………………………………………………………..18
2.3.2 Breakpoint Processes………………………………………………18
2.3.3 Biological Processes………………………………………………..19
2.3.4 Ion Exchange Processes…………………………………………….21
2.3.5 Membrane Processes……………………………………………….22
2.3.6 Aeration Processes…………………………………………………23
2.4 Microbial Processes in Ammonia Cycling, Oxidation, and Reduction…….…24
2.4.1 Microbial Oxidation………………………………………………..24
2.4.2 Microbial Reduction………………………………………………..26
3. Materials and Methods
3.1 Pilot Plant Design and Operation…………………………………………….28
3.2 Water Chemistry……………………………………………………………..32
3.3 Microbial Culturing…………………………………………………………..35
5 3.4 Isolation of Nucleic Acids……………………………………………………37
3.5 amoA and 16S Gene Library Construction and Phylogenetic Analysis…...... 38
4. Results
4.1 Manuscript I…………………………………………………………..……..41
4.2 Manuscript II…………………………………………………………..…….64
5. Discussion…………………………………………………………………….…...….93
6. Future Work…………………………………………………………………...…..…94
7. References……………………………………………..………………………….….95
Appendix………………………………………………………………………………....99
6 1. INTRODUCTION
1.1. BACKGROUND
With the exception of mixing ammonia-cleaning agents with bleach, ammonia is
+ rarely associated with human health problems. Ammonia in its ionized form NH4 is
called ammonium and is not toxic to aquatic life whereas in its un-ionized form NH3, it is
extremely toxic to aquatic life. As the primary ionization state of ammonia is as
ammonium, its presence in pretreated water and finished drinking water is not regarded as a contaminant by the United States Environmental Protection Agency (US EPA).
Though the compound proper of ammonia does not call for alarm, it is in the chemistry of ammonia where problems arise in drinking water treatment. Ammonia has
the ability to provide organisms with a great source of energy as it is oxidizable by a
group of prokaryotes residing within the !-Proteobacteria and Chrenarcheota. It is the
oxidized forms of ammonia that threaten human health. In pre-treated drinking water
these forms of oxidized ammonia, nitrite and nitrate, have a maximum contaminate limit
(MCL) set by the US EPA at 1 mg.L and 10 mg/L, respectively. When ingested, nitrate is
reduced to nitrite and binds oxyhemoglobin converting it to methylhemoglobin.
Methylhemoglobin is not capable of carrying oxygen and can be lethal. Blue baby
syndrome is a common concern in areas with high nitrite/nitrate levels in drinking water.
In addition to direct impact on human health, ammonia has indirect impact as
well. As ammonia is primarily a groundwater contaminant, its presence is usually accompanied by iron and arsenic from natural mineral deposits in the aquifer. The physicochemical removal of arsenic can be coupled concurrently to the iron removal
7 process, a common and cost effective method for iron removal. The primary oxidization state of arsenic in ground water requires it to be oxidized to sorb to oxidized iron particles for subsequent removal by filtration. The presence of ammonia in source waters can interfere with this process by chemically consuming the added oxidant required to oxidize the arsenic, potentially exposing consumers to arsenic levels above the US EPA
MCL of 10 ug/L.
The consumption of chlorine increases the chlorine demand of the system leading to a decrease in the free chlorine residual in the distribution system. The lower disinfectant residual may allow biological growth in the distribution system and lead to nitrification events and disease outbreaks.
It is therefore beneficial to remove ammonia in source waters prior to treatment.
Technologies that exist for ammonia removal include ion exchange, air stripping, membrane processes, biological processes, and breakpoint chlorination. These technologies all have strengths and weaknesses, but of particular interest are biological processes as their potential benefits outweigh the weaknesses compared to other technologies. Biological processes have been used in Europe and wastewater industry for years, but have only recently been considered in the United States. The slow acceptance in the U.S. is primarily due to negative perceptions related to microorganisms as well as a poor understanding of the complex microbial community and its role related to filter reliability.
Herein, I attempt to elucidate the microbial community of a full-scale biologically active filter, demonstrate its effectiveness at oxidizing ammonia, as well as test its reliability and ease of implementation at the pilot-scale.
8 1.2. OBJECTIVES
As ammonia in source waters leads to several operational and distribution system
issues in the drinking water industry, it is my hypothesis that a rapid sand filter can
become biologically active. After an initial seeding event a period of time will be required to establish a population of microorganisms capable of complete nitrification. These organisms responsible for nitrification will not be negatively affected by changes in operational parameters, the addition of arsenic, and filter backwashing. It is my intent to show that biologically active filters for ammonia oxidation are indeed an economical, reliable, and sustainable technology that could greatly reduce several problems associated with drinking water distribution. I will also attempt to resolve the work of Lytle et al.
(2007) who concluded a full-scale filter oxidized ammonia and arsenic via biological processes.
1.3. APPROAC
Ammonia oxidation was monitored in two pilot-scale rapid sand filters located at
the USEPA Andrew W. Breindbach Environmental Research Center (AWBERC) in
Cincinnati, OH. Microbial analysis consisted of standard microbial culturing techniques
to monitor effluent and influent waters for heterotropic plate counts (HPC), total
coliforms, E. coli., enterococci, and aerobic endospore forming bacteria. Chemical measurements were used to assess filter reliability in response to changes in operation parameters such as backwash and water chemistry.
9 Molecular analysis was performed on a full-scale drinking water filter to determine
the microbial consortia responsible for ammonia oxidation. 16S and amoA clone libraries
were constructed from nucleic acid extracts of filter media and subject to sequence and
phylogenic analysis.
1.4 OVERVIEW OF THESIS
This thesis is broken down into seven chapters, which include the introduction, literature review, materials and methods, results, discussion, and future work.
Chapter 1 contains a short introduction, the study objectives, and an overview of the thesis.
Chapter 2 contains a brief literature review and background information on ammonia and treatment methods. Section 2.1 contains an overview of arsenic behavior in
groundwater, including the chemistry, distribution in the environment, and transport.
Section 2.2 discusses ammonia health effects, including chronic and acute, carcinogenic and non-carcinogenic, and beneficial effects. Section 2.3 contains an overview of available ammonia treatment technologies, including, breakpoint processes biological processes, ion exchange, membrane processes, and aeration processes. Section 2.4 delves further into microbial processes related to ammonia (i.e., microbial ammonia reduction,
oxidation and concludes with a brief discussion of applications (i.e., reactor studies).
Chapter 3 describes the materials and methods used in the study. Section 3.1 describes the filter media and water samples collected from the full-scale plant and the setup and design of the pilot-scale columns. Section 3.2 describes the methods used for water chemistry analysis. Section 3.3 describes the methods used for microbial culturing
10 of indicator organisms in the pilot-scale system. Section 3.4 describes the method used to isolate DNA from media and water samples. Section 3.5 describes the procedure used in the amoA and 16S-rDNA gene phylogenetic analysis.
Chapter 4 contains the results. Section 4.1 describes the results of DNA isolation and clone library analysis of media removed from the full-scale filter. Section 4.2 contains the results of the pilot-scale filter operation.
Chapter 5 provides a discussion of these findings and the relevance to the body of literature on microbial treatment processes.
Chapter 6 provides some ideas for future work on this topic.
Chapter 7 lists general references.
An Appendix containing additional information is provided at the end of the document.
2. LITERATURE REVIEW
2.1 OVERVIEW OF AMMONIA BEHAVIOR IN WATER
Ammonia in aqueous solution may exist as complexed compounds, ionized
ammonia, or in a dissolved gaseous state. The chemical state ammonia is a critical factor
for its removal from water. This section discuses the formation of these chemical states
in water.
2.1.1. AMMONIA CHEMISTRY
The US EPA does not regulate ammonia in source waters, nor is it regulated at
any point in the distribution system. As such, understanding ammonia chemistry in water
11 is essential as it reacts with water but also chlorine, the primary disinfectant added to
many distribution systems. In pure water, ammonia can be in a “free” unionized state
+ (NH3) or in an ionized state called ammonium (NH4 ). These two forms exist in an
equilibrium that is a function of the water pH and temperature (Anthonisen, 1976). Free
ammonia is gaseous and is therefore capable of leaving solution whereas ionized
ammonia remains soluble. Free ammonia in water reacts with hydronium ions to produce
ammonium, and the resultant ammonium is a strong conjugate acid, which will react with any base to form free ammonia. The equilibrium in an aqueous solution is dependent on the concentration of the hydromium ion and thus the solutions pH. As the pH is lowered, via the addition of hydronium ions, the equilibrium shifts to the right forming ammonium. As hydroxide is consumed by hydronium, ammonium forms due to the lack of base to remove a proton from ammonium.
Ammonia binds free chlorine used for disinfection on the order of 7.6 mg of Cl2
per mg of ammonia (Andersson, 2001). The combination of ammonia and chlorine forms
conpounds called chloramines. Chloramines are typically used as a disinfectant in
drinking water distribution systems when there is potential of disinfection by-product
(DPB) formation using chlorine. The oxidizing power of chloramines is less than that of
chlorine such that fewer halogenated DBPs are formed. The use of chloramines as a
disinfectant requires longer contact times for disinfection of some bacteria and virus.
With use of chloramines as disinfectant, and their inherent lower oxidative power, stable insoluble passivating films on the surface of lead pipes can become soluble, raising lead levels above the EPA action level of 15 ppb (Zhang, 2008 and Zhang, 2009).
Chloramines may break down in the distributions system releasing free ammonia (Jafvert
12 and Valentine, 1992). With the ability to break down, utilities utilizing chloramines
disinfection are essentially feeding a sequestered substrate into the distribution system. If
ammonia oxidizing bacteria (AOB) are present, a nitrification event may take place
subjecting consumers to nitrite and nitrate levels above the US EPA Maximum
Contaminant Limit (MCL).
Chloramine formation in source water can be a problem if arsenic removal is
coupled with the iron removal process (Lytle, 2007). Specifically, the consumption of
the chemical oxidant (chlorine), by excess free ammonia, used to oxidize arsenic (III) to
arsenic (V) will prevent arsenic (V) formation and lead to a detrimental effect on arsenic
removal via sorption to iron particles (Lytle, 2007).
2.1.2. AMMONIA DISTRIBUTION
Primary natural sources of ammonia include decay and excretion by plants and
animals. Though these sources are important on the biosphere scale, their relative
importance to groundwater contamination is minimal due to low local concentrations.
Such levels are thought to be degraded by soil biota before reaching groundwater aquifers deep beneath the surface.
Of greater importance to groundwater is anthropogenic contamination. Figure 1 shows the application of nitrogen as fertilizer in the United States. Though total nitrogen is presented, the majority of nitrogen in fertilizer is in the form of ammonia or ammonium.
As soil biota can degrade a portion of this applied nitrogen, concentrations are so high that seepage into groundwater aquifers is common.
13 Another anthropogenic source of ammonia is from livestock farms. Untreated animal waste can accumulate rapidly and pollute local watersheds and shallow aquifers.
Figure 2 is a map of USGS water quality assessments for ammonia in surface and ground waters. It is clear that the mid-western United States faces the greatest levels of free ammonia due to agriculture practices and subsequent fertilizer runoff. The figure demonstrates the wide scope of excess ammonia in source waters and the applicability of the technology described herein.
Existing with ammonia in groundwater sources are elements such as arsenic, iron, and manganese. With the US EPA’s MCL for arsenic at 10 ppb and the interferences caused by ammonia in physicochemical iron/arsenic removal, it would prove to be beneficial if ammonia could be removed from source waters using an affordable, sustainable, and readily implemented technology.
14
Figure 1. Nitrogen Fertilizer Loading in the United States.
15
Figure 2. USGS National Water Quality Assessment of Filtered Ammonia in Source waters.
2.2 AMMONIA HEALTH EFFECTS
The US EPA does not regulate ammonia in the source water or distribution
system due to its low toxicity profile. However, products of ammonia oxidization are
toxic. Nitrite and nitrate are regulated in the source water at a MCL of 1 and 10 mg/L,
respectively.
2.2.1. Carcinogenic Effects and Risks
There is no evidence that ammonia is carcinogenic and it is not classified as such
by the US EPA, Department of Health and Human Services (DHHS), or the (IARC). (US
DHHS, 2004; US EPA 1989)
16
2.2.2. Non-Carcinogenic Effects and Risks
2.2.2.1. ACUTE EXPOSURE
In animal tests, oral exposure of ammonium salts yielded an LD50 of 350-750 mg/kg of body weight (INRS, 1987; Smyth, 1941). Single doses of ammonium salts in the range of 200-500 mg/kg of body weight did cause acidosis, kidney damage, and nervous system dysfunction. In humans, ammonia becomes toxic when it is ingested at a rate greater than detoxification. At ammonium salt levels between 75-360 mg/kg of body weight, the anion of the salt may be of greater consequence as it induces acidosis. Some research has suggested that a physiological adaptation may be able to overcome the induced acidosis (Geneva, 1986). Ammonium chloride, at levels greater than 100 mg/kg of body weight, has been implicated in reducing tissue sensitivity of insulin by shifting the acid-base equilibrium (US EPA, 1989). Female rats given a single dose of 290 mg of ammonia per kg of body weight per day inhibited fetal weight compared to controls
(Geneva, 1986).
2.2.2.2. CHRONIC EXPOSURE
Male rats given drinking water containing 478 mg of ammonium per kg of body weight per day significantly lowered body weight, bone mass, calcium and blood pH.
The animals also had lower body weight and lower fat accumulation than controls
(Geneva, 1986). In female rats, ammonia can cross the placental barrier and, if in large enough quantity, decrease egg production (US DHHS, 2004). No literature was found on
17 chronic exposure of humans to ammonia. Campbell et al. (1958) did report that the
human taste threshold is 34 mg/L, which would correspond to 0.97 mg/kg/day if 2 L of
water were consumed per day by an individual with an average body weight of 70 kg.
2.2.3. BENEFICIAL EFFECTS
No literature was found on beneficial effects of ammonia. There is anecdotal evidence that the consumption of ammonium salt, specifically ammonium chloride, (also known as Sal-ammoniac) helps neuralgia of the face as well as catarrh and bronchitis
(Family Physician, 1886).
2.3 Treatment Technology Options
2.3.1. OVERVIEW
There are many physicochemical processes available for the removal of ammonia
from water, however many of these processes are applied only to wastewater and are
required to meet the US EPA Clean Water Act for nitrogen discharge into environmental
waters. Many of the technologies discussed below require significant energy input and/or
high operational costs. Biological process may be the exception with full-scale plants
requiring little to no infrastructure change, capital investment, and operational attention.
2.3.2. BREAKPOINT PROCESS
To achieve breakpoint chlorination one must routinely exceed 5 mg/L chlorine,
far more than what is required for disinfection, to obtain a free chlorine residual.
Andersson et al. (2001) calculated the theoretical chlorine consumption as 7.6 mg of
18 chlorine per milligram of ammonia. The overdosing is required to overcome the chlorine
demand caused by free ammonia. The chlorine demand is defined as the difference
between the amount of chlorine added and the free, combined, or total chlorine residual at the end of the chlorine contact period. The chlorine demand in water increases with high concentrations of dissolved and suspended organic compounds as well as inorganic
reducing agents such as ferrous iron, nitrite, and others. In treatment of excess ammonia,
the breakpoint technique requires the addition of chlorine to the process water until a point is reached such that all free ammonia, and oxidizable organics and inorganics, are oxidized by chlorine (Pressley, 1972; Westerhoff, 2004; Cotruvo, 1981; Fleischacker,
1983; Moore, 1951).
As this method is routinely used in the drinking water industry to control excess ammonia and nitrification in distribution systems, it is difficult to calculate the correct dose of chlorine to reach the breakpoint. This is made difficult due to changes in the water quality of source waters. In the case of ammonia-nitrogen levels exceeding 1 mg/L, utilities must accurately calculate and dose the chlorine to ammonia-nitrogen. This ratio is important in that it defines the type of chloramine residual in the distribution system after the breakpoint process (Pressley, 1972).
2.3.3. BIOLOGICAL PROCESSES
Biological treatment has the ability to overcome the disadvantages of
conventional physicochemical processes but it is not with its own drawbacks. The
European Council Directive on the quality of water has set the maximum threshold for
ammonia at 0.5 mg/L (EC 1998b). As such, European water utilities have found
19 biological processes to be economical, efficient, and compatible for the oxidation of
ammonia coupled with other physicochemical processes for contaminant removal.
Research has shown that ammonia oxidation works well between 10 and 23oC. Kihn et
al. (2002) reported that oxidation of 1.2 mg/L and 0.4 mg/L of ammonia was inhibited at
temperatures below 5oC at a loading rate of 5 mh-1 with an empty bed contact time
(EBCT) of 3.6 min. However, they found that the temperature rise in spring allowed rapid recovery of oxidation. Andersson et al. (2001), fed 1.2 mg/L ammonia and observed 40-90% oxidation in pilot-scale filters at temperatures greater than 10oC, and
greater than 90% oxidation in full-scale filters receiving an average ammonia influent of
20-120 mg/L. Oxidation in both full- and pilot-scale filters decreased to 10-40% oxidation at temperatures between 4 and 10oC.
With the utilization of biological processes, sufficient filter colonization time is required before the biological process operates at capacity. As previously discussed,
AOB are known for their slow growth rates compared to other species thus, startup times on the order of 40-60 days have been reported for maximum ammonia oxidation (Lytle
2007, Rittmann and Snoeyink 1984, Andersson 2001).
Bacterial slough off from hydraulic shear may breakthrough the filter and contaminate the distribution system with heterotrophic and pathogenic bacteria if they are present in the biological filter. It was recommended by Rittmann and Snoeyink (1984) that a physical or chemical barrier should separate the biological filter effluent from the distribution system. To minimize cost and increase efficiency, biologically active filters may be included at the beginning of the treatment train and followed by conventional physicochemical processes and finally chlorination to control biological breakthrough.
20 The carbon to nitrogen (C:N) ratio is important to know when choosing biological
treatment in order to avoid lithotroph/heterotroph competition. The optimal C:N ratio is
dependent on a number of factors as detailed by Verhagen and Laanbroek (1991).
Lithotroph/heterotroph competition may reduce the desired biological activity and cause
biofouling of the filter.
With biological nitrification allowed to occur in the treatment plant under
relatively controlled conditions, issues arising from the drop in pH and alkalinity, slough-
off of bacteria, and decrease in Dissolved Oxygen (DO) can be addressed prior to the
water entering the distribution system. Rittmann and Snoeyink (1984) reported that
nitrification as pretreatment developed ‘biologically stable’ water that resisted biological regrowth in the distribution system due to the lack of biodegradable organic/inorganic
material (BDO/IM). As a secondary effect of BDO/IM removal, chlorine demand and
formation of DBPs are lower.
2.3.4. ION EXCHANGE PROCESSES
Ion exchange processes utilize the exchange of anions or cations between two electrolytes. One electrolyte is typically in the aqueous form while the other is bound to an ionic substrate such as a zeolite, resin or mineral. Ion exchange has the advantage that it is a re-generable technology. Once the resin has exhausted its ability to bind its substrate, the resin is backwashed with a brine solution to replace the bound ion with an exchange ion that typically has lower affinity. Ions with less affinity to the resin than the bound ion displace the bound ion via a competitive binding process. Since the exchange ion affinity is in higher quantity, the bound ion is exchanged due the higher number of exchange ions. Ion exchange has the disadvantage of creating a secondary waste stream
21 in that the backwash brine solution contains high concentrations of bound ions. This
waste stream forces the utility to manage disposal, which can potentially be hazardous and costly. The second disadvantage is that ammonia must be in the ionized form
ammonium. As presented earlier, ionization of ammonia requires protonation of the un-
charged ammonia molecule and thus strict pH control of the water (Du, 2004). This pH
adjustment must be maintained by the utility, and can prove difficult with varying
influent water characteristics.
Using a natural clinoptilolite zeolite in a packed bed column and an influent of 20
mg/L ammonium, Demir et al. (2002) observed 1 mg/L ammonia breakthrough after 450
bed volumes (BV) using a superficial velocity of 25 and 50 BV/h. Breakthrough was
observed after 300 BV when using a flow rate of 75 BV/h.
2.3.5. MEMBRANE PROCESSES
Many studies have been conducted to determine membrane technology’s ability to remove ammonia, but full-scale technology is limited to medium and large systems due
to cost constraints. Several studies reported the removal rate using reverse osmosis (RO)
membrane technology range from 60-90% when ammonia salts are the primary form of
ammonia. Awadalla et al. (1994) observed greater than 99% removal using RO
membranes and 66% when using nanofiltration (NF), in mine effluent water. Synthetic
waters containing only ammonium exhibited lower removal efficiencies in both NR and
RO (10-30%) indicating that ammonia-salt complexes are required for high removal
efficiency. Kurama et al. (2002) observed high removal of ammonia (96.9%) using
cellulose-acetate membranes and synthetic water containing ammonia salts. In real
water, removal efficiencies were negatively affected by increased salinity. Nevertheless,
22 membrane systems have several potential advantages such as the ability to modularize, low equipment footprint, no chemical additives, automatic operation, no DO limitations, and few temperature effects. Membrane technology is limited to large scale and newly designed utilities, due to the specialized control setups required and large capital investment in terms of installation, operation, and maintenance. Assuming a 3 year membrane life, Koyuncu et al. (2001) suggest that $0.95-$1.06/m3 of water compared to the average using conventional treatment of $0.4/m3.
2.3.6. AERATION PROCESSES
Mostly applied to wastewater treatment, this process involves mass transfer across an interface. In the case of water, the interface of interest is the air-water interface.
As ammonia is a very soluble compound and is involved in fast equilibrium with ammonium, removal by air stripping is complicated. Ammonium must be converted to the volatile (un-ionized) ammonia by raising the pH of the water above 10. Though pH adjustment is sufficient for conversion of the bulk ammonium to ammonia, temperature plays an important role (Norddahl, 2006). For example, Norddahl et al. (2006) found that increasing the temperature by 5oC in water with a pH greater than 10 decreased the mass transfer coefficient, shortening the mass transfer rate across the interface. With a dependence on temperature, this technology may not be suited for areas subject to source water temperature fluctuations. Implementation of such a technology also requires major infrastructure change and capital investment on the part of the utility along with precise engineering control over the process. This process has the advantage of eliminating the
23 waste stream associated with ion exchange brines as well as completely removing nitrogen
from the system.
2.4 MICROBIAL PROCESSES IN AMMONIA CYCLING,
OXIDATION, AND REDUCTION
With the diverse metabolism exhibited by microorganisms, the ability exists to
both oxidize and reduce inorganic nitrogenous compounds including oxides and salts. This
section will provide an overview of the basic principles involved in microbial oxidation
(nitrification) and reduction (denitrification). Both of these processes have, in some way,
been exploited in water or wastewater treatment.
2.4.1. MICROBIAL OXIDATION
Bacteria responsible for ammonia oxidation are chemolitoautotrophs with growth
rates on the order of 26-60 hours (Rimon and Shilo, 1982; Belser, 1984). These bacteria
use inorganic substrates for electrons donors and inorganic carbon for incorporation into
biomass. The bulk of ammonia oxidation is carried out aerobically with bacteria using
oxygen as the terminal electron acceptor. Biological oxidation of ammonia (NH4) to
- nitrate (NO3 ) involves a two-step sequence of reactions. Bacteria in the genera
Nitrosomonas, Nitrosococcus, Nitrosospira, Nitrosolobus, and Nitrosovibiro are
responsible for the first step of the nitrification reaction (Hagopian and Riley, 1998) as
follows:
+ - + NH4 + 1.5 O2 ! NO2 + H2O + 2H (1)
Nitrobacter, Nitrococcus, Nitrospira, and Nitrospina are responsible for the second step
(Watson and Waterbury, 1971) as follows:
24 - - NO2 + 0.5 O2 ! NO3 (2)
By summing these equations, the overall nitrification reaction is obtained:
+ - + NH4 + 2 O2 ! NO3 + 2 H + H2O (3)
These equations are net reactions involving a complex series of enzyme-catalyzed intermediate steps.
Only two enzymes are required for oxidation of ammonia to nitrite; a cytoplasmic trans-membrane protein, ammonia monooxygenase (AMO) and periplasmic protein hydroxylamine oxidoreductase (HAO) (Hollocher, 1982; Yamanaka and Sakano, 2007).
Nitrite oxidoreductase (NOR) is the sole enzyme responsible for nitrite oxidation to nitrate, (equation 2) (Hussain-Allem and Sewell, 1981).
The first step of ammonia oxidation catalyzed by AMO is endergonic with net energy being gained only after oxidation of hydroxylamine to nitrite by HAO. The decrease in pH that accompanies ammonia oxidation is in the second step where nitrous acid is formed, thus, when utilizing this as a treatment technology, it is important to maintain sufficient alkalinity to avoid feedback inhibition by hydrogen (Siegrist and
Gujer, 1987).
The oxygen demand of nitrification is also significant. For complete nitrification,
+ 4.57 mg O2 is required per mg NH4 - N oxidized (U.S. EPA 1975). Application in drinking water treatment would require aeration of ground water to prevent DO limitations and ammonia breakthrough. With maximum DO saturation of water, ammonia oxidation is limited to approximately 1.7 mg/L. free ammonia. US EPA has piloted methods to continuously aerate the biological filter to remove free ammonia above 1.7 mg/L (unpublished data).
25 Heterotrophs may oxidize organically bound ammonia, but their contribution to
total nitrification in the environment is believed to be negligible. This process is believed
to be co-metabolism rather than an energy-generating pathway (Isirimah, 1976; Witzel and Overbeck, 1979; Papen 1989).
Archea of the Chrenarchoeta have recently been reported to play a significant role in ammonia oxidation, though physiological tests are needed to determine substrate affinity and kinetics (Francis, 2005).
Ammonia oxidizing bacteria are fairly hardy and can remain viable in the absence of substrate for prolonged periods, to be rapidly activated after introduction of utilizable substrate (Kumar and Nicholas, 1983). Chemolithotropic bacteria responsible for ammonia oxidation have also been found to oxidize urea and methane as well as co- metabolize hydrocarbons (Ward, 1987; Bedard and Knowles, 1989).
It is imperative to know the microbial community of biologically active filters if the technology is to be implemented. The individuals responsible for degradation of contaminants can be identified and further water amendments to support their growth and metabolism may be added. Knowing the community may also provide information on operational reliability, depending on factors such as mass transfer kinetics of individual organisms, biofilm structure, susceptibility to disinfection, and filter modeling.
Identifying pathogenic organisms in the filter is crucial to safety and public perception issues of relying on “biological incubators” for treatment of drinking water.
2.4.2. MICROBIAL REDUCTION
The converse of microbial ammonia oxidation is the reduction of intermediates of nitrification. Denitrification may remove nitrogen from the system by way of molecular
26 nitrogen formation or merely reduce nitrate and nitrites. Denitrification requires a limited
oxygen (anoxic) environment for the use of nitric oxides as electron acceptors. Nitrate is the primary electron acceptor with nitrite, nitric oxide, and nitrous oxide less thermodynamically favored (Lin, 1998). Denitrification may occur in soils, marine sediments, and groundwater. Soils undergo denitrification primarily during periods of water saturation as oxygen’s ability to diffuse is limited by the water (Knowles, 1982).
Organisms responsible for denitrification are diverse and can be found in all phylogenic lineages and include primarily heterotrophic bacteria. A diverse flora, rather than a single species, similar to nitrification, accomplishes complete denitrification
(Knowles, 1982).
- - NO3 ! NO2 ! NO + N2O ! N2 (g) (1)
Denitrification may also occur in anaerobic environments by way of ammonia
oxidation to molecular nitrogen. This process is called Anaerobic Ammonia Oxidation
(Anammox). In Anammox, ammonia is converted to nitrogen gas using nitrite as an
electron acceptor (Eq 2). Anammox organisms grow by utilizing carbon dioxide as a
carbon source and nitrite as the electron donor. This process is mediated by members of
Planctomycetes (Ahn, 2006).
+ - NH4 + NO2 ! N2 + 2H2O (2)
Though not a denitrification process, in term of nitrogen removal, nitrate may be
reduced to ammonium by a process known as dissimilatory nitrate reduction to
ammonium (DNRA). As a means of nitrate reduction in the environment, this process is
less common than denitrification and requires organisms carrying the nrf gene (Smith,
2007; Simon, 2002).
27
3 MATERIALS AND METHODS
3.1. PILOT PLANT DESIGN AND OPERATION
The pilot scale filters were setup and operated at the US EPA AWBERC in
Cincinnati OH. Two clear Polyvinyl Chloride (PVC) pipe sections, 2.5 inches in
diameter, were cut to 72 inches in length and fitted at the top with two schedule 40 (S40)
socket tees to allow overflow (Figure 1, Section 4.2). The bottom of the columns
consisted of a 2.5 inch S40 socket-to-socket union with a 2.5 inch socket to ! inch male pipe thread (MPT) PVC reducer bushing. A ! inch stainless steel (SS) MPT to " inch
MPT fitting was screwed into the reducer bushing. A S40 PVC female pipe thread (FPT)
to socket fitting was fixed and glued to a 3-way socket-socket-socket valve using a short
piece of " inch S40 PVC. All PVC to PCV connections were cleaned using primer
cleaner and glued using all-purpose cement. One side of the 3-way valve was reduced to
# inch using a S40 PVC socket to FPT fitting fixed to the 3-way valve with a short piece
of " S40 PVC. A " SS MPT to # inch MPT reducer compression fitting was inserted
and connected to # inch tee. The tee was oriented such that the middle male connector faced up. The middle connector was attached to a # inch SS ball valve and then 72 inches of # inch Teflon tubing that paralleled the column for headloss measurements.
The other # compression fitting was attached to a length of Teflon tubing and run to a
Cole-Parmer 25210-51 gear pump fitted with a Cole-Parmer 81492-398 pump head. The gear pump was controlled using Cole-Parmer 75211-46 controller. The effluent side of the gear pump was attached to a length of Teflon tubing and up parallel to the column and fixed to the waste manifold at the top, behind the column. The remaining socket on
28 the 3-way valve was connected to ! inch S40 PVC and run to a 1/8 hp pump used for backwashing. The backwash pump was connected to a 40L carboy using ! inch S40
PVC. The 40L carboy served as a reservoir for the storage of water used for backwashing. The columns were mounted vertically onto a unistrut frame using two 2.5 inch tube clamps at the top and bottom of the column. Headloss lines and tape measures were mounted parallel to the column using unistrut clamps. An influent manifold was constructed using " inch SS fittings as well as " inch Teflon tubing. The columns were filled with a combination of sand, gravel, and/or anthracite coal. The columns were subsequently wrapped in aluminum foil to protect organisms from light and to simulate real world conditions.
Feed system
The columns were initially fed using two 4000 L SS basins. The basins were situated approximately 4 feet above the top of the columns on the second floor of the pilot plant. The basins were connected using 1.5 inch S40 PVC and fitted with PVC ball valves to allow concurrent or independent filling/draining. At a point sufficient to allow concurrent or independent operation of the basins, a S40 socket-1/2 inch FPT-socket tee was fitted. A ! inch SS MPT to 3/8 inch MPT compression fitting was inserted and connected to a length of 3/8 inch Teflon tubing running to 1/16th HP pump. The output
of the pump was connected to a length of 3/8 Teflon tubing to a 50 L SS drum fitted with
an overflow 2 inches from the top rim of the tank and an output valve at the bottom. This
50 L tank served as a constant head tank (CHT). A 3/8 inch SS ball valve was attached to
the output of the tank and then to a 3/8 inch needle valve leading to the bottom of a
29 floating body flow meter reading 0-4 L/min. The top of the flow meter was attached to a
3/8 inch Teflon tube and run to the influent manifold at the top of the columns. The columns were filled with tap water dechlorinated using two Pentek C1 inline carbon cartridge filters.
Water Dosing
Initially the columns were fed using deionized (DI) water containing sufficient solid ammonium chloride (Fisher #A661) and sodium arsenite (Mallinckrodt #7392) to give a final concentration of 1 mg/L and 100 µg/L, respectively and mixed for 5 minutes using industrial mixers. At day 38 the feed water was changed to dechlorintaed municipal tap water. The water was dechlorintaed using Pentek C1 inline carbon cartridge filters. Water storage basins were filled at a rate sufficient to achieve dechlorination and filters discarded after chlorine breakthrough (2 filling cycles). The pH was adjusted with each dosing to pH 8.0 +/- 0.2 using 6 M H2SO4, typically this required approximately 25 mL of acid.
Re-configuration of Feed system:
At day 212 the pilot columns were moved to another location in the AWBERC building due to construction as a result, the feed system had to be re-configured. The chemical dosing of the basins and separate CHT was eliminated through by the instillation of in-line chemical feed ports on the effluent of the new water basin. A 700 L
SS basin was used as a CHT and storage basin for dechlorinated water. The tank was fitted with a floating valve approximately 4 inches from the top of the basin. The floating
30 valve allowed the basin to fill at a rate equal to the effluent rate but shutoff in the event of
a hydraulic backup. The float was attached to the effluent port of a series of Pentek C1 carbon filters that fed the system with dechlorinated tap water. A ! inch SS ball valve was connected to the bottom of the tank and fitted to a ! inch needle valve then the
bottom of a floating body flow meter reading 0-4 L/min. The top of the flow meter was
connected to ! inch S40 PVC consisting of three ! inch socket-1/2 inch MPT-1/2 inch
socket PVC tees and two ! inch PVC in-line static mixers downstream of the three tees.
Nylon ! inch MPT-1/8 hose barb fittings were screwed into the ! inch MPT S20 PVC
tee and served as injection ports. The end of the last static mixer was attached to ! inch
S40 PVC and run to the influent manifold at the top of the columns on the floor beneath.
The new setup was again fed via gravity. All other aspects of the system remained identical to the original setup. Continuously mixed chemical solutions were fed from 40
L HDPE carboys using a multi-head peristaltic pump using size 14 tubing connected to separate nylon hose barb injection ports.
*All PVC to PVC connections were cleaned using purple primer cleaner and glued using
all-purpose cement, and all pipe thread connected used Teflon tape.
Column operation:
The columns were initially operated at 300 mL/min through the filter, measured
weekly using a graduated cylinder and stopwatch. This initial flow rate equated to a
hydraulic loading rate (Equation 1) of 2.5 gpm/sqft and an empty bed contact time of 10
minutes (Equation 2).
Flow rate (gpm) LR (gpm ft 2 ) = (Equation 1) Area of Filter bed (ft 2 )
31
Volume of filter media (ml) EBCT (minutes) = (Equation 2) flow rate (ml/min)
When a change in hydraulic loading rate was made, equation 1 was used to find the required flow rate to increase the loading rate by the desired unit of loading rate.
Once this change was made, the effluent flow rate was measured and confirmed using a graduated cylinder and stopwatch.
Columns were backwashed based on a set headloss of 30 inches or more.
Backwash was performed using a flow rate sufficient to give 50% bed depth expansion for 10 minutes. Light backwashes were performed in the event of column dryout. In this chase, the column was backwashed lightly (less than 50% expansion) for no longer than 5 minutes. This proved to be sufficient to remove air pockets and eliminate water channeling.
3.2. WATER CHEMISTRY
Chemical samples were collected from the common influent and unique effluent lines of each column. Every effort was made to collect daily chemical samples, but circumstances existed such that this goal was not always attainable. Table 3.2.1 lists the sampling and monitoring strategy and Table 3.2.2 lists the sampling and analytical procedures. QA/QC procedures and standard methods can be found in the appendix. The pH and dissolved oxygen (DO) was measured with a Hach Company (Loveland, CO)
EC40 bench-top pH/DO meter (Model 50125). The instrument was standardized for pH
32 and DO daily using a two-point calibration with pH 7 and 10 standard solutions and
saturated tap water, respectively (Whatman, Hillsboro, OR). All metals except arsenic
were measured with a Thermo Jarrel Ash (Franklin, MA) 61E® purged inductively coupled argon plasma spectrometer (ICAP) or ThermoFisher 6000 ICAP. Ammonia-
nitrogen, nitrate-nitrogen, and nitrate-nitrogen were measured using standard method
350.1 and 353.2. Total organic carbon was measured using standard method 5310C.
High-performance liquid chromatography/ Inductively coupled plasma mass
spectrometry (HPLC/ ICP-MS) was conducted at the University of Cincinnati following
internal laboratory QA/QC protocols.
Chromatographic separations were completed with an Agilent 1100 liquid
chromatograph (Agilent Technologies, Palo Alto, California) equipped with binary HPLC
pump, an autosampler, a 100 !L sample loop, and a vacuum de-gasser system. Reversed
phase chromatography was performed with a ZORBAX EclipseXDB-C18 column (5mm
"4.6mm id"250mm) from Agilent Technologies (Santa Clara, CA).
Tetrabutylammonium hydroxide was chosen as the ion-pairing reagent as it provided the
resolution needed for the separation. For the counter ion, hydroxide was chosen, as it did
not produce baseline interference that other counter ions containing Cl would.
The ICP-MS used was an Agilent 7500ce (Agilent Technologies, Toyoko, Japan).
It was operated using standard resolution mode and programmed to read m/z 75 for
arsenic.
33 Table 3.2.1. Sampling/Monitoring Strategy
Sample/Monitoring Total No. Matrix Measurement Frequency Experimental QC Location Samples
Concurrent with Ammonia, Nitrite, Hydrant/Filter water Automated Matrix Spikes, Lab hydrant or filter Colorimetric and Nitrate Blanks operation
TOC Concurrent with Hydrant/Filter water Combustion hydrant or filter (total organic Lab Blank
operation carbon)
Hydrant/Filter Spring and DNA concentrate PCR Lab Blank Autumn
Determined by
Concurrent with experimental Metal & Non-metal Extracted hydrant/filter ICAP hydrant or filter Matrix Spikes design ions water operation
Enterococci
Total Coliforms Concurrent with
Hydrant/filter influent, E. Coli hydrant or filter Cultures Lab Blank backwash/ effluent, backwash AOB
operation SRB
Endospores
Automated Matrix Spikes, Lab Kinetics Filter concentrate Once per week Colorimetric Blanks
34 Table 3.2.2. Sampling/Analytical Procedures
Sampling/ Holding Time Container Analysis Sample Quantity Measurement Sample Preservation (hours) Method
Cultures 1 L PP <30 SOP 3 Cool at 4 ºC DNA 8-10 L sample bottle 72
0.90ml Concentrated HDPE sample ICAP 60 mL sample water Extracted volume Nitric acid <366 bottle pH <2
Inorganic analysis 250 mL sample HDPE sample (ammonia, nitrite, Extracted volume Cool at 4 ºC <72 water bottle nitrate)
1 drop of
concentrated H2PO4 <336 TOC 40 mL sample water Extracted volume Glass sample vial pH<2
Cool at 4 ºC
Kinetics 100L sample water Concentrated volume HDPE carboy N/A <24
3.3. MICROBIAL CULTURING
Several types of standard microbial measurements were made in addition to
molecular methods. All sample vessels used in microbal analysis were made of
polypropylene washed with DI water and autoclaved at 121oC for a minimum of 15
minutes. Some sample vessels contained sodium thiosulfate.
E. coli was cultured following standard method 9223 using the Colilert method by
35 IDEXX. 100 ml of sample was added to a pre-dispensed media for the multiple-tube
method. 100 ml of sample was added to the quantitray-2000 sleeve and processed
through a lamination device to seal the sleeve. The sleeve was then placed in a 35oC incubator for 24 hours. A positive was recorded as fluorescence under 366 nm UV light.
Positive samples are reported as most probable number (MPN) per 100 ml of sample.
Total Colifoms were cultured following standard method 9223 using the Colilert method by IDEXX. 100 ml of sample is added to a pre-dispensed media for the multiple- tube method. 100 ml of sample are added to the quantitray-2000 sleeve and processed through a lamination device to seal the sleeve. The sleeve is then placed in a 35oC
incubator for 24 hours. A positive is recorded as a color change from colorless to yellow.
Positive samples are reported as most probable number (MPN) per 100ml of sample
(SM9221C).
Enterococci was cultured following ASTM #D6503-99 using the Enterlert method
by IDEXX. 100 ml of sample is added to a pre-dispensed media for the multiple-tube
method. 100 ml of sample are added to the quantitray-2000 sleeve and processed through
a lamination device to seal the sleeve. The sleeve is then placed in a 41oC incubator for
24 hours. A positive is recorded as fluorescence under 366 nm UV light. Positive samples are reported as most probable number (MPN) per 100 ml of sample (SM9221C).
Heterotrophic Plate Counts allow the enumeration of live cultureable heterotrophs from potable water on R2A agar. Per standard method 9215C, counts are made based on
presence or absence of bacterial colonies on plates grown at 28oC for 7 days. Serial dilutions were made using 9 ml dilution blanks consisting of Butterfield’s buffer (pH
7.0+/-0.2). Dilutions were carried out using 3 dilution blanks per sample. For example:
36 1 mL of sample was added to a 9 mL blank and mixed. This gave a 10:1 dilution. This procedure was carried out two more times. 100 µl of each dilution was spread onto an
R2A plate. For example 100 µl of the 10:1 dilution spread on a plate gave a final dilution of 10-2. Colony-forming units (CFU) describe the number of viable colonies that have grown on the media and therefore present in the sample. Data is reported as CFU/ml of sample plated. Duplicate plates of each dilution were made for each sample tested and average counts were calculated and recorded.
Aerobic Endospores can be used to evaluate treatment unit processes including physical removal (coagulation and clarification), disinfection, and to determine the order of magnitude removal efficiencies from surface water sources. Endospores are also used for determining the efficacy of processes such as halogen, ozone, and ultraviolet inactivation. The presence of endospores may be used to determine the physical integrity of drinking water distribution systems that may have been compromised by pipeline breaks or maintenance procedures Per standard method 9218, counts are made based on the presence or absence of bacterial colonies on a filter membrane placed on nutrient agar and incubated at 35oC for 24 hours. A wide-field dissecting microscope is used to count
bacterial colonies that have arisen from endospores. These colonies are reported as
CFU/ml of sample plated.
3.4. ISOLATION OF NUCLEIC ACIDS
Microbes in water samples were concentrated by filtration prior to DNA
extraction, in order to increase DNA yield. Typically, 100 ml of water sample was
collected and immediately filtered onto a 0.22 µm filter (Pall #4806). The filter unit was
37 capped and stored at -20oC until DNA extraction. Filter units were thawed to room
temperature and aseptically cut into quarters using a scalpel and petri dish. Three
quarters of the filter was placed in a sterile tube and stored at -80oC. The remaining quarter filter was placed into a sterile 2ml microcentrifuge tube containing glass beads. 2
µl of 50 µg/ml Proteinase K (EpiCentre Biosciences) and 400 µl of tissue and cell lysis
buffer (EpiCentre Biosciences) was added to each tube. The tubes were incubated at
65oC for 10 minutes and vortexed every 3 minutes. The tubes were then placed in a MP bioscience bead-beater and beadbeated for 40 seconds on setting 6. Tubes were removed
and placed on ice for 5 minutes then centrifuged for 10 minutes at >10,000 rpm. The
supernatant was transferred to a clean 2 ml centrifuge tube and 350 µl of MPC protein precipitation solution (EpiCentre Biosciences) was added to 600 µl of lysed extract. The
sample was vortexed for 10 seconds and centrifuged for 10 minutes at >10,000 rpm. The supernatant was carefully removed and placed into 500 µl of ice cold isopropanol (Acros organics), then centrifuged for 10 minutes at >10,000 rpm. The isopropanol was decanted and the pellet was washed twice with ice cold 75% ethanol (Sigma). The ethanol was decanted and residual ethanol allowed to evaporate in a biological cabinet.
Samples were resuspended in 50 µl of sterile water (Chemicon TMS-006-B) and stored at
-20 oC.
3.5. AMOA AND 16S GENE LIBRARY CONSTRUCTION AND PHYLOGENETIC
ANALYSIS
All PCR reactions were performed using a BioRAD Tetrad thermal cycler. Table
3.5.1 lists primers and references to reaction cycle programs used for this project.
38 Table 3.5.1. Primers and References to Reaction Conditions.
Primer Sequence (5’ to 3’) Reference Amplicon (bp)
AroA-USA-F GTSGGBTGYGGMTAYCABGYCTA Inskeep et al. ~500 AroA-USA-R TTGTASGCBGGNCGRTTRTGRAT 2007
AMOA-F GGGGTTTCTACTGGTGGT Hoshino et al. ~500 AMOA-R CCCCTCRGSAAAGCCTTCTTC 2001
16s-F (11F) GTTTGATCCTGGCTCAG Kelly et al. 2005 ~1500 16s-R (1512AR) ACGGYTACCTTGTTACGACTT
T3 ATTAACCCTCACTAAAGGGA Invitrogen ~1600 T7 TAATACGACTCACTATAGGG Cloning Kit
PCR products were electrophoresed on a 1.8% agarose gel in 0.5X TAE buffer for approximately 50 minutes. 4 µl of 10,000X GelStar nucleic acid stain was added to 60ml of agarose dissolved in 0.5X TAE. A molecular weight standard was included in each set of lanes to confirm amplicon size. The gel was visualized and documented using a
Kodak GelLogick UV imaging system. Bands used for cloning were excised from the gel using a sterile scalpel taking care to reduce excessive agarose and nonspecific bands.
The gel slices were then processed using a gel extraction kit (Qiagen #28604). Gel extracted DNA was eluted in 10 µl of sterile water and 1 µl of DNA subject to electrophoresis following the aforementioned method to verify quantity and recovery after purification. Once a determination on quantity was made, based on comparison to known molecular weight standards of similar size, the appropriate volume of gel extracted DNA was added to a TOPO TA cloning kit (Invitrogen #450030) ligation reaction to obtain a vector to insert molar ratio of 1:1. The TOTO TA cloning reaction
39 was performed following manufactuarers protocol except in the case of 16s amplicons.
Due to the size of this amplicon a ligation time of 15 minutes was used. Transformed cell
were plated in 30 µl and 70 µl aliquots onto LB agar plates containing 100 µg/ml
ampicillin and incubated at 32oC overnight. Colonies that grew overnight on LB agar were picked with sterile toothpicks and cultured at 32oC in 96 well plates containing LB broth supplemented with 100 µg/mL ampicillin overnight in a rotary shaker at 125 rpm.
Using a 12 channel pipettor, 2 µl of the overnight cultures in LB broth were added to 12
tube strips of PCR tubes containing PCR master mix and T3/T7 primers. These were amplified according to Table 3.5.2. After electrophoresis, PCR tubes that contained the appropriate (~1400 bp) amplicon were pipetted to a new 96 well plate and stored for sequencing. This procedure was followed for all gene libraries except amoA ligations
were incubated for 5 minutes.
Table 3.5.2. PCR Reaction Setup.
PCR Reaction Protocol
Concentration of Final Per RXN Reagent Units Setup Stock Concentration (µL)
Buffer 10.0 X 0.60 1.50 1.5
DNTP 2.5 mM (each) 0.20 2.00 2.0
11F 25.0 µM 0.30 0.30 0.3
1512AR 25.0 µM 0.30 0.30 0.3
Taq 5.0 U/µL 0.20 0.20 0.2
H2O ------19.70 19.7
DNA 25.0 ng/µL 25.00 1.00 --
Sequencing reactions were performed on an ABI 3730 using the BigDye
terminator system. Sequences were edited in Mega4 and compared to sequences in the
40 NCBI database using the BLAST function. Highest similarity sequences were downloaded for each sequence. Duplicate sequences were discarded. Known sequences of identified species were also downloaded and aligned.
With similar sequences downloaded from the NCBI database, phylogenetic trees were constructed. Outgroups were selected from previous literature. Nucleic acid sequences were aligned and phylogeny constructed using the neighbor joining method.
2000 bootstrap replications were performed.
4. RESULTS
4.1 MANUSCRIPT I:
Microbial Characterization of a Full-Scale Biologically Active Filter for Drinking Water Treatment
By
Colin P. White1, Andrea F. Galloway2, Ronald W. DeBry1, and Darren A. Lytle3*
1University of Cincinnati, Department of Biological Sciences Cincinnati, OH 45221
2University of Cincinnati, Department of Biomedical Engineering Cincinnati, OH 45221
3United States Environmental Protection Agency, ORD, NRMRL, WSWRD, TTEB, Cincinnati, Ohio 45268
ABSTRACT
Biological nitrification has been used as a reliable technology in wastewater treatment for decades. Implementing biological approaches to drinking water treatment has faced resistance in the United States due in part to the lack of understanding of microbial processes and concerns of biostability of microbial populations in
41 biologically active filters. The objective of this study was to apply culture and non-culture based biological methods to identify the microbial community of a full scale biologically active drinking water filter reported to have ammonia and arsenic oxidation capabilities. Results showed that 1.13 mg/L of ammonia nitrogen was consistently oxidized to nitrate without any effects of seasonal variation. The major ammonia oxidizers were identified as Nitrosomonas based on amoA clone libraries. In addition to several dozen genera, 16S clone libraries confirmed that Nitrosomonas was the dominant ammonia oxidizer and Nitrospira was the dominant nitrite oxidizer. Members of Sphingomonas and Rhizobiales were also found to dominate the clone libraries. Together, these results shed light on the microbial ecology of biologically active filters and show that biologically treatment of drinking water is both reliable and safe.
42 INTRODUCTION
Biologically active filtration is commonly used in Europe and Asia for drinking
water treatment (Bouwer and Crowe, 1988). Biological processes have the potential to
cut operation costs by decreasing the amount of chemicals required for treatment and
increasing efficiency in terms of decreased biological re-growth in the distribution system
(DS) and decreased chlorine demand (LeChevallier, 1992; Snoeyink, 1984; Camper,
2003). However, biological processes have not been widely accepted in the United
States, mainly due to issues arising from the negative perception of microorganisms, as
well as questionable reliability and effectiveness (LeChevallier, 1992; Andersson, 2001;
Chung, 2006). With the lack of historical operational data using biological processes,
and the complexity of microbial systems, there exists a need to both document
biologically active systems and design experimental systems to elucidate the microbial
consortia and the effects of operational parameters.
One use of biologically active filters for drinking water treatment involves the
regulation of nitrate/nitrite nitrogen levels. Excessive ingestion of nitrite and nitrate can be hazardous (Shuval, 1974), so the United States Environmental Protection Agency (US
EPA) has set a source water maximum contaminant level (MCL) for nitrite and nitrate at
1 mg/L and 10 mg/L respectively. Yet, no MCL exists for ammonia. As utilities are only required to monitor for these contaminants in the source water, their concentrations may build in the distribution system via uncontrolled partial- or full-nitrification. When
excessive levels of free ammonia are present in the source water or ammonia is added to
form chloramines, nitrification may occur with sufficient dissolved oxygen (DO)
(Skadsen, 1993). Nitrification in the DS, and the pH drop associated with nitrification,
43 can impact the corrosion rates of distribution system and premise materials (Murphy,
1997; Edwards, 2008). In addition, the increased chlorine demand and growth of heterotrophic biofilms produce undesirable taste and odor issues (Burlingame and
Anselme, 1995). Excess ammonia itself may interfere (by way of chloramine formation)
with the maintenance of a free chlorine residual in the distribution system and the
chemical oxidation of arsenic (III) in treatment plants utilizing the iron removal process for arsenic removal (Lytle, 2007a: 2007b).
Lytle et al (2007a; 2007b) reported on the use of biologically active filters to oxidize ammonia and arsenic in a full-scale water treatment plant. They demonstrated that the filters completely and consistently oxidized 1.13 mg/L of ammonia to nitrate and
30 µg/L of arsenic (III) to arsenic (V) without addition of a chemical oxidant.
Preliminary filter analysis and follow-up pilot studies identified bacteria as the source of ammonia and arsenic oxidation (Green, 2007). The findings of Green (2007) and Lytle
(2007b) were based primarily on culture-dependent methods, with little use of culture- independent (molecular) methods. Molecular microbiological techniques to characterize microbial communities in wastewater processes and tricking filters, characterization of full-scale biologically active drinking water treatment systems is limited (Kihn, 2000;
Park, 2006; Qin, 2007; Van der Wielen, 2009). There is a clear need to better identify the diversity of bacterial communities, including the presence of human pathogens, in biologically active drinking water filters to improve the understanding of such a complex system. Identification of the microbial consortia will provide a greater insight into the dynamics of biologically active filters, establish their susceptibility to pathogen growth, and determine their applicability as a viable treatment technology.
44 The goal of this study is to identify members of the microbial community in a
full-scale drinking water filter and identify the specific microbes responsible for
ammonia oxidation. Specifically, we re-visit the microbial oxidation system previously
characterized by Lytle et al (2007a; 2007b).
EXPERIMENTAL PROCEDURES
Water quality analysis. The pH was measured on-site with a Hach Company
(Loveland, CO) EC40 benchtop pH/ISE meter (Model 50125) and a Hach Company
(Loveland, CO) combination pH electrode (Model 48600) with temperature corrections.
A two-point calibration with pH 7 and 10 standard solutions standardized the instrument
daily (Whatman Hillsboro, OR). Dissolved oxygen (DO) was measured with a Hach
Company (Loveland, CO) Model DO175 DO meter and a Model 50180 DO probe. Free
and total chlorine were measured with a Hach DR/2000 spectrophotometer (Loveland,
CO) using the DPD method (Standard Method, 4500-Cl G). A Thermo-Fisher (Waltham,
MA) 6000 purged inductively coupled argon plasma atomic emission spectrophotometer
(ICP-AES) was used to analyze total iron and other elements including Mn, S, Ca, Mg,
Si, P, etc. (U.S. EPA Method 200.7) and an inductively coupled argon plasma atomic mass spectrophotometer (ICP-MS) (U.S. EPA Method 200.8) was used to measure arsenic and other elements. In the field, ferrous iron was measured using the 1,10 phenanthroline method (APHA-AWWA-WEF, 2005). Total iron was measured by the same method with the addition of a reducing reagent in the reagent powder pillow provided by the Hach Company to convert Fe(III) to Fe(II).
45 Ammonia (NH3) was measured in the field with a Hach DR890 colorimeter using
method 66. Additionally, nitrate was measured by a Hach DR890 colorimeter using
methods 53 and 54. Nitrate and nitrite were further measured by ion chromatography
(U.S. EPA Method 300.0) in the laboratory. Nitrate samples were preserved at 4oC and
kept at pH<2 via H2SO4 and were analyzed within 28 days. The nitrite was analyzed
within 48 hours.
Other water samples were also sent back to either the U.S. EPA or Battelle
laboratories for analysis. Dissolved inorganic carbon (DIC) was analyzed by a
coulometric procedure on a UIC Model 5011 CO2 coulometer (Joliet, IL) with Model 50
acidification module, operated under computer control. Syringe filters (0.2 µm) (Anotop
25®, Whatman, Inc., Clifton, NJ) were used to separate colloidal iron from soluble iron.
The biological counts of the water samples, filter backwash solids, and filter
media were assayed within hours of sampling. In the case of solids, biological analysis
was performed on the excess water collected with the solids after the sample was briefly
mixed. Total heterotrophic plate counts (HPCs) were determined by Standard Method
9215 C using R2A media and incubation at 22°C for seven days. Ammonia-oxidizing bacteria (AOB) were enumerated based upon nitrite/nitrate detection following 30 days
incubation at 28°C in Soriano-Walker media using most probable number (MPN)
estimates and using 10 tubes/dilution (APHA-AWWA-WEF, 2005).
Electron microscopy. A small amount of filter media was fixed in a pH 7.2 cacodylate-buffered 1% paraformaldehyde 2.5% glutaraldehyde mixture. The media was post-fixed in 1% OsO4 and dehydrated in an ethanol series, dehydrated in
Hexamethyldisilazane (HMDS) for 1 hour, and then air dried in a desiccator. The media
46 was coated with gold/palladium prior to scanning electron microscopy (SEM). Samples were viewed using a JEOL 6490LV (Peabody, MA) SEM at 30 kV under high vacuum.
Energy dispersive x-ray analysis (EDXA) was performed for 60 live seconds using a process time of 4 and working distance of 10 mm.
DNA extraction. The media was collected during filter backwash using aseptic techniques. Samples were taken to the US EPA laboratories in Cincinnati and processed immediately. 1.2 g of media was placed in 400 µl of lysis buffer (EpiCentre Biosciences) containing SDS and sonicated three times for 30 s each, vortexing between steps. The tubes were then centrifuged, and the supernatant was removed and placed into a tube containing glass beads. The supernatant was beaten with the beads for 1 minute. 2 µl of
50 µg/mL Proteinase K (EpiCentre Biosciences) was added to each tube and incubated at
65°C for 10 minutes. The supernatant was extracted and nucleic acids precipitated in ice cold isopropanol at 4°C for 30 minutes. The nucleic acids were pelleted and washed with
70% ice cold ethanol and desiccated. Samples were rehydrated in 50 µl of sterile water
(Chemicon TMS-006-B) and stored at -20°C.
PCR, cloning, and sequencing. PCR was conducted in a 25 µl volume containing 0.2 µM DNTPs, 0.3 µM each primer, 0.4U Taq polymerase (Takara), 0.6X
Taq buffer and 1 µl of template DNA. amoA PCR contained 0.1 µg/µl of nonacetylated
BSA. PCR cycling conditions are described in Table 2. PCR products were electrophoresed on a 1.8% agarose gel in 0.5X TAE buffer. PCR bands used for cloning were excised from the gel using a sterile scalpel and processed using a gel extraction kit
(Qiagen #28604). Products were cloned using the TOPO TA cloning kit (Invitrogen
K4575-01).
47 Sequencing reactions were performed on an ABI 3730 using the BigDye
terminator system with T3/T7 primers. Sequences were edited and aligned in MEGA4 and compared to sequences in the NCBI database using the BLAST function. Highest similarity sequences were downloaded for each sequence. Duplicate and chimeric sequences were discarded. Chimeric sequences were identified using Mallard
(Ashelford, 2006). Type-cultured and previously identified sequences of bacterial species with close BLAST hits were retrieved from GenBank.
Phylogenetic analysis. Phylogenetic trees were constructed using MEGA version 4 (Tamura, Dudley, Nei, and Kumar 2007). Phylogeny was inferred using the neighbor-joining algorithm with 2000 bootstrap replicates assuming pairwise deletion using Maximum Composite Likelihood distance correction.
RESULTS
Water quality. Detailed water quality analysis can be found in previous studies by this lab (Lytle et al, 2007a; 2007b). Table 1 summarizes the water quality of the treatment plant. Raw and influent ammonia values averaged 1.13 mg/L with nitrate and nitrite below the detection limit. After filtration, prior to the addition of chlorine,
ammonia was oxidized to below 0.03 mg/L with nitrite and nitrate measuring 0.02 mg/L
and 1.3 mg/L respectively. Ammonia oxidizing bacteria (AOB) counts were highest on
the filter media and lowest in the plant effluent. HPC were highest in well samples and
lowest in the plant effluent (Table 2).
Electron microscopy. SEM-EDX analysis of filter media showed particles
coated with an outer layer of iron and manganese (Figure 1a). No biofilm was
48 observed on the outermost layer on of the media on any of the particles examined.
Differential backscatter imaging identified media particles lacking the outer coating, presumably from scouring during backwash or SEM processing. These particles resolved a complex biofilm, containing a mix of spirochete, bacilli, and cocci within a thick extracellular matrix (Figure 1b). The observation of biofilm under the outer layer is consistent with AOB counts (Table 1).
Molecular methods. DNA isolations from filter media provided sufficient template for PCR. PCR products of 16S and amoA genes were the correct length and of suitable yield for cloning. A total of 431 16S and 61 amoA clones were selected and sequenced. After removing identical duplicate and chimeric sequences, 297 and 31 unique 16S and amoA sequences were grouped into phylotypes (OTUs), using 97% sequence identity as a threshold value, and subsequently analyzed for phylogeny.
Representatives from each OTU were used as query sequences to NCBI BLAST to identify close relatives. These sequences were then downloaded and included in their respective phylogeny.
amoA libraries. Unique amoA sequences were grouped into 9 OTUs, 4 of which were singletons. 8 OTUs were clustered within the genus Nitrosomonas with the remaining single OTU closely resembling Nitrosospira (Figure 2). Within Nitrosomonas,
7 OTUs, comprising 28 total sequences, fell within the N. oligotropha lineage. The remaining OTU, comprising 2 sequences, fell within the N. europaea lineage. Primers directed to the archaeal-amoA gene were used in an attempt to create a clone library for
49 subsequent identification of AOA, but PCR did not produce any amplification after exhaustive optimization efforts.
16S libraries. Unique 16S sequences were grouped into 65 OTUs, 36 of which were singletons (Figure 3). With nearly 55% of OTUs singletons, diversity was driven by species captured only once, suggesting a highly diverse system. The 65 OTUs were classified into 9 discrete groups (Table 3). 27 OTUs (together accounting for 21% of all sequences) were closely related to Alpha-proteobacteria. Within the Alpha- proteobacteria, the orders Rickettsia, Rhodobacterales, Rhizobiales, and
Sphingomonadales were represented. Beta-proteobacteria accounted for 15% of sequences with 10 OTUs representing Nitrosomonadales, Methylophilales, and
Burkholderiales. Gamma/Delta/Epsilon-proteobacteria were grouped together, and accounted for approximately 3% of the sequences within 8 OTUs. Analysis of 19 sequences representing 6 OTUs identified only unknown bacteria as close relatives.
These sequences were placed into the unknown group. Planctomycetes, Bacteroidetes,
Verrucomicrobia, and Chloroflexi accounted for approximately 4% of sequences within 9
OTUs.
The phylum Nitrospirae, known to be nitrite oxidizers (Skadsen, 1993; Schramm,
1999), dominated the clone library with over 51% of sequences grouped into 5 OTUs.
This result strongly supports the notion that members of Nitrospirae, rather than members of the order Rhizobiales (Nitrobacter), are the dominant nitrite oxidizing bacteria in the nitrification process in this filter, perhaps due to low nitrite and DO concentrations
(Manser, 2005).
50 One OTU (16 sequences) was identified as “Candidatus Nitrotoga arctica”, a
cold adapted nitrite oxidizing bacterium previously isolated from activated sludge
(Alawi, 2007).
DISCUSSION
Lytle et al (2007a) observed that raw ground water contained an average of 1.13
mg/L of ammonia-nitrogen prior to filtration and less than 0.1 mg/L after filtration.
Nitrate- and nitrite-nitrogen levels prior to filtration were below detection. After
filtration, nitrate-nitrogen was present stoichiometrically with pre-filtration ammonia nitrogen. This stoichiometric relationship and presence of nitrate-nitrogen after filtration suggested that ammonia was oxidized in the filters. As no means of oxidation was included in the treatment train, biological nitrification was determined to be the causative agent. Therefore, the filters were investigated for their microbial diversity and correlated to biological ammonia oxidation.
HPCs of raw and effluent plant water showed that the 0.87 mg/L free chlorine residual in the plant effluent was sufficient to destroy any bacteria carryover from the filters. Backwash HPCs were highest, suggesting that the heterotrophic populations in the filters do not form strong biofilms (Table 2). AOB-MPN counts support prior studies that observed nitrification can be attributed to microorganisms in the filters (Table 2).
Furthermore, AOB-MPN counts were highest in sonicated media samples, as opposed to backwash samples, suggesting that the AOB form biofilms on the filter media and that this biofilm is strong enough to protect the organisms from shear forces associated with elevated water flow during the backwash process.
51 Agreement of SEM observations and AOB counts suggests that the biofilm
formed prior to the deposition of solids during the filtration process and that the outer layer is porous to allow diffusion of reduced arsenic and free ammonia to the bacteria.
This outer layer may protect the biofilm from shear forces associated with water flow.
This possible reduction in shear may account for the rapid recovery of biological activity
after backwash (Celmer, 2008).
In relation to 16S and amoA clone libraries, general congruence was observed
with respect to the AOB. The dominant nitrifying organisms present were members of
the phyla Proteobacteria and Nitrospirae. Of the sequences in the amoA library, species
from the genus Nitrosomonas were found to be the major ammonia oxidizers, accounting
for greater than 98% of total sequences. Over 95% of total sequences clustered in the N.
oligotropha lineage, and 3% in the N. europaea lineage. One sequence was related to the
genus Nitrosospira. These findings are consistent with other research suggesting
Nitrosomonas, specifically the species N. oligotropha, are better adapted to low ammonia
concentrations than other AOB and dominate the drinking water distribution system
(Regan, 2002; 2003).
The filters operate under relatively limited ammonia concentrations, and thus
limited nitrite concentrations, so Nitrospira are expected to be the dominant nitrite
oxidizers due to their lower half-saturation coefficient for oxygen and nitrite compared to
Nitrobacter (Schramm, 1999). Though many clones were related to the order
Rhizobiales, no sequences were related to the genus Nitrobacter. Though Nitrobacter
has a higher growth rate, its ability to compete for oxygen and substrate is lower than that of Nitrospira (Martiny, 2005). Moreover, the detection of “Candidatus Nitrotoga artica”
52 may give beneficial operational flexibility to the filter. These organisms have been
shown to oxidize nitrite at temperatures near 4oC, possibly adding to the operational
temperature range (Alawi, 2007). Rhizobiales, though not typically related to aquatic
environments, may be introduced by external sources or play a novel role in the nutrient limiting environment (Schmeisser et al, 2003).
Recent attention has been focused on the dominance of AOA in the environment
(van der Wielen, 2009; Francis, 2005). Specifically, many studies have shown that archaea are the dominant ammonia oxidizers in both soil and marine environments. To this end, archaeal-amoA primers were used to detect their presence and serve a cloning insert for gene library construction and sequence analysis. No AOA were detected based on the absence of an archaeal-amoA amplicon. The lack of detection of AOA indicates that AOB dominate the system. This may be due to the fact that the raw water may contain compounds that inhibit the growth or physiology of these organisms or that the
ammonia levels in the raw water saturate the AOA’s ability to oxidize the ammonia, thus
selecting for AOB.
Sphingomonas and Rhizobiales dominated the 16S clone library. Sphingomonas
has been shown to degrade complex organic molecules such as xenobiotics, chlor/nitro
phenolics, and large polymers. They accomplish this via numerous dependent and
independent metabolisms that may add to the operational flexibility of biologically active
filters (White, 1996; Nohynek; 1995; Kawai, 1999; Fredrickson, 1995; Balkwill, 1997).
Members of Rhizobiales have demonstrated the ability to utilize a broad range of carbon sources under aerobic conditions (Pang, 2007). Studies on pure cultures of
Rhizobiales have shown that they may also be capable of degrading methyl parathion,
53 metolachlor, polyacrylamides and quaternary ammonium alcohols, all potential source water contaminates (Kaech; Qiu 2006; Padden, 1997; Yang and Lee, 2008).
An interesting finding of this study was the fact that no known pathogenic bacteria were identified as the majority of the 16S clone libraries. A primary concern of biologically active filtration is whether or not the filter is hospitable for pathogenic organisms, so this finding is encouraging. Though such organisms may be sensitive to chlorination, the possibility exists for distribution system contamination via slough-off if there is a malfunction in chlorination or the organism is capable of forming endospores.
Therefore, there exists a need to further study this question in greater detail.
Electron micrographs show distinct diverse cellular morphology. Of note, a spirochete morphology strongly resembling members from the genus Leptospira was observed (Figure 1c) (Johnson and Faine, 1984). Primers from the literature were used in an attempt to amplify the flaB gene from Leptospira for subsequent cloning and identification. Preliminary evidence for molecular detection is positive, though more work needs to be completed to determine if it is pathogenic or saprophytic.
The metabolic diversity known to exist in organisms identified in biologically active filters may provide the only means to remove complex contaminates from source waters. Such filters may also serve as a unique source for isolation of novel organisms that may be beneficial for bioremediation.
ACKNOWLEDGEMENTS
The authors would like to acknowledge fellow U.S. EPA staff Daniel Williams and Christy Muhlen for their field assistance, Keith Kelty, Brittany Almassalkhi, and Bill
54 Kayler for analytical support and Amy Reed, Clifford Johnson, and Laura Boczek for microbiological analysis support. We would also like to thank Mike Watts from Greene
County Sanitary Engineering Department. Finally, we would like to thank Kyle Hawkins
of Miami University (Ohio) for editorial comments. Any opinions expressed in this paper are those of the authors and do not necessarily reflect the official position and policies of the U.S. EPA. Any mention of products or trade names does not constitute recommendation for use by the U.S. EPA.
55 REFERENCES
1. Alawi, M., A. Lipski, T. Sanders, E.-M. Pfeiffer, and E. Spieck. 2007. Cultivation of a novel cold-adapted nitrite oxidizing beta proteobacterium from the Siberian Arctic. ISME Journal. 1:256–264.
2. Andersson, A., P. Laurent, A. Kihn, M. Prevost, and P. Servais. 2001. Impact of temperature on nitrification in biological activated carbon (BAC) filters used for drinking water treatment. Water Res. 35:2923-2934.
3. Ashelford, K.E. 2006. New Screening software shows that most recent large 16S rRNA gene clone libraries contain chimeras. Appl. Environ. Microbiol. 72:5734- 5741
4. Balkwill, D. L., G. R. Drake, R. H. Reeves, J. K. Fredrickson, D. C. White, D. B. Ringelberg, D. P. Chandler, M. F. Romine, D. W. Kennedy, and C. M. Spadoni. 1997. Taxonomic study of aromatic-degrading bacteria from deep- terrestrial-subsurface sediments and description of Sphingomonas aromaticivorans sp nov, Sphingomonas subterranea sp nov, and Sphingomonas stygia sp nov. Int. J. System. Bacteriol. 47:191–201.
5. Bouwer, E. J., and P. B. Crowe. 1988. Biological processes in drinking water treatment. J. Am. Water Works Assoc. 80:82-93.
6. Camper, A. K., K. Brastrup, A. Sandvig, J. Clement, C. Spencer, and A. J. Capuzzi. 2003. Effect of distribution system materials on bacterial regrowth. J. Am. Water Works Assoc. 95:107-121.
7. Celmer, D., J. A. Oleszkiewicz, and N. Cicek. 2008. Impact of shear force on the biofilm structure and performance of a membrane biofilm reactor for tertiary hydrogen-driven denitrification of municipal wastewater. Water Res. 42:3057- 3065.
8. Chung, J., H. Shim, S. J. Park, S. J. Kim, and W. Ba. 2006. Optimization of free ammonia concentration for nitrite accumulation in shortcut biological nitrogen removal process. Bioproc. Biosyst. Eng. 28:275–282.
9. Clesceri, L. S., A. E. Greenberg, and A. D. Eaton. 1998. Standard Methods for the Examination of Water and Wastewater, 20th ed. American Public Health Association; Am. Water Works Assoc. and Water Environ. Fed.: Washington, DC.
10. Francis, C. A., K. J. Roberts, M. J. Beman, A. E. Santoro, and B. B. Oakley. 2005. Ubiquity and diversity of ammonia-oxidizing archaea in water columns and sediments of the ocean. PNAS. 102:14683-4688.
56 11. Fredrickson, J. K., D. L. Balkwill, G. R. Drake, M. F. Romine, D. B. Ringelberg, and D. C. White. 1995. Aromatic-degrading Sphingomonas isolates from the deep subsurface. Appl. Environ. Microbiol. 61:1917–1922.
12. Green, C. N. 2007. M.S. thesis. University of Texas, Austin, TX.
13. Hill, C. B., and E. Khan. 2008. Comparative study of immobilized nitrifying and co-immobilized nitrifying and denitrifying bacteria for ammonia removal from sludge digester supernatant. Water, Air and Soil Pollution. 195:23-33.
14. Johnson, R. C., and S. Faine. 1984. Leptospira. p. 62–67. In N. R. Krieg and J. G.Holt (ed.), Bergey’s manual of systematic bacteriology, vol. 1. Williams & Williams, Baltimore, MD.
15. Kaech, A., N. Vallotton, and T. Egli. 2006. Isolation and characterization of heterotrophic bacteria able to grow aerobically with quaternary ammonium alcohols as sole source of carbon and nitrogen. System. Appl. Microbiol. 28:230- 41.
16. Kawai, F. 1999. Sphingomonads involved in the biodegradation of xenobiotic polymers. J. Ind. Micro. Biotech. 23:400-407.
17. Kihn, A., P. Laurent, and P. Servais. 2000. Measurement of potential activity of fixed nitrifying bacteria in biological filters used in drinking water production. J. Inds. Micro. Biotech. 24:161-166.
18. LeChevallier, M. W., W. C. Becker, P. Schorr, and R. G. Lee. 1992. Evaluating the performance of biologically active rapid filters. J. Am. Water Works Assoc. 84:136-146.
19. Lytle, D. A., T. J. Sorg, L. Wang, C. Muhlen, M. Rahrig, and K. French. 2007. Biological nitrification in full-scale and pilot-scale iron removal drinking water treatment plant filters. Jour. Water Supply: Res. and Tech.: AQUA. 56:125- 136.
20. Lytle, D. A., A. S. Chen, T. J. Sorg, S. Phillips, and K. French. 2007. Biological As(III) oxidation in water treatment plant filters. J. Am. Water Works Assoc. 99:72-86.
21. Manser, R., W. Gujer, and H. Siegrist. 2005. Consequences of mass transfer effects on the kinetics of nitrifiers. Water Res. 39:4633-4642.
22. Martiny A. C., H. J. Albrechtsen, E. Arvin, and S. Molin. 2005. Identification of bacteria in biofilm and bulk water samples from a nonchlorinated model drinking water distribution system: detection of a largenitrite-oxidizing population associated with Nitrospira spp. Appl. Environ Microbiol. 71:8611-8617.
57
23. Nohynek, L. J., E. L. Suhonen, E. L. Nurmiaho-Lassila, J. Hantula, and M. Salkinoja-Salonen. 1995. Description of four entachlorphenoldegrading bacterial strains as Sphingomonas chlorophenolica sp nov. System. Appl. Microbiol. 18:527–538.
24. Padden, A. N., F. A. Rainey, D. P. Kelly, and A. P. Wood. 1997. Xanthobacter tagetidis sp. nov., an organism associated with Tagetes species and able to grow on substituted thiophenes. Int. J. Syst. Bacteriol. 47:394-401.
25. Pang, C. M., and W.-T. Liu. 2007. Community structure analysis of reverse osmosis membrane biofilms and the significance of Rhizobiales bacteria in biofouling. Environ. Sci. Technol. 41:4728-4734.
26. Park, H. D., G. F. Wells, H. Bae, C. S. Criddle, and C. A. Francis. 2006. Occurrence of ammonia-oxidizing archaea in wastewater treatment plant bioreactors. Appl. Environ. Microbiol. 72:5643-5647.
27. Qin, Y. Y., D. T. Li,, and H. Yang. 2007. Investigation of total bacterial and ammonia-oxidizing bacterial community composition in a full-scale aerated submerged biofilm reactor for drinking water pretreatment in China. FEMS Microbiol Lett. 268:126-134.
28. Qiu, X.-H., W.-Q. Bai, Q.-Z. Zhong, M. Li, F.-Q. He, and B.-T. Li. 2006. Isolation and characterization of a bacterial strain of the genus Ochrobactrum with methyl parathion mineralizing activity. J. Appl. Micro. 101:986-994.
29. Regan, J. M., G. W. Harrington, H. Baribeau, R. De Leon, and D. R. Noguera. 2003. Diversity of nitrifying bacteria in full-scale chloraminated distribution systems. Water Res. 37:197-205.
30. Regan, J. M., G. W. Harrington, and D. R. Noguera. 2002. Ammonia- and nitrite-oxidizing bacterial communities in a pilot-scale chloraminated drinking water distribution system. Appl. Environ. Microbiol. 68:73-81.
31. Rittmann, B. E., and V. L. Snoeyink. 1984. Achieving biologically stable drinking water. J. Am. Water Works Assoc. 76:106-114.
32. Schmeisser, C., C. Stöckigt, C. Raasch, J. Wingender, K. N. Timmis, D. F. Wenderoth, H. C. Flemming, H. Liesegang, R. A. Schmitz, K. E. Jaeger, and W. R. Streit. 2003. Metagenome survey of biofilms in drinking-water networks. Appl. Environ. Microbiol. 69:7298-309.
33. Schramm, A., D. de Beer, J. C. van den Heuvel, S. Ottengraf, and R. Amann. 1999. Microscale distribution of populations and activities of Nitrosospira and Nitrospira spp. along a macroscale gradient in a nitrifying bioreactor:
58 quantification by in situ hybridization and the use of microsensors. Appl. Environ. Microbiol. 65:3690-3696.
34. Skadsen, J. 1993. Nitrification in a distribution system. J. Am. Water Works Assoc. 85:95–103.
35. White, D. C., S. D. Sutton, and D. B. Ringelberg. 1996. The genus Sphingomonas: physiology and ecology. Curr. Opinion. Biotech. 7:301-306.
36. Witzig, R., W. Manz, S. Rosenbergerb, U. Krügerb, M. Kraumeb, and U. Szewzyk. 2002. Microbiological aspects of a bioreactor with submerged membranes for aerobic treatment of municipal wastewater. Water Res. 36:394- 402.
37. van der Wielen, P. W., S. Voost, and D. van der Kooij. 2009. Ammonia- oxidizing bacteria and archaea in groundwater treatment and drinking water distribution systems. Appl. Environ. Microbiol. 75:4687-4695.
38. Yang, C. F., and C. M. Lee. 2008. Pentachlorophenol contaminated groundwater bioremediation using immobilized Sphingomonas cells inoculation in the bioreactor system. J. Hazard. Mater. 152:159-65.
39. Zhang, Y., A. Griffin, and M. Edwards. 2008. Nitrification in premise plumbing: Role of phosphate, pH and pipe corrosion. Environ. Sci. Technol. 42:4280–4284.
40. Shuval, H.I., and N. Gruener. 1974. Effects on man and animals of ingesting nitrates and nitrites in water and food. In: Effects of Agricultural Production on Nitrates in Food and Water with Particular Reference to Isotope Studies, IAEA, Vienna, Austria STI/PUB361 .
59 Table 1. Results of Water Quality Analysis
Before Filtration After Filtration After Chlorine Analyte Raw (BF) (AF) Addition (AC) pH 7.48± 0.1 7.76± 0.2 7.76± 0.1 7.68± 0.0
Alkalinity (mg/L as CaCO 3) 373± 11 376± 12 365± 7 360± 10 DO (mg/L) 1.22± 0.4 6.74± 0.4 5.4± 1.0 6.33± 0.6 ORP (mV) _114± 19 36± 57 187± 144 598± 36 Temperature (ºC) 14± 2 14± 2 14± 2 16± 2 Free Chlorine (mg/L) NA 0.03± 0.0 0.04± 0.2 0.87± 0.3 Total Chlorine (mg/L) NA 0.04± 0.1 0.03± 0.2 0.97± 0.3 Ammonia (lab) (mg/L as N) 1.15± 0.0 1.11± 0.0 0.1± 0.1 <0.1 Nitrate (mg/L as N) <0.04 <0.04 1.11± 0.2 1.15± 0.1 Nitrite (mg/L as N) <0.01 <0.01 0.02± 0.0 <0.01 TOC (mg/L) 1.2 1.1 1.1 1 Total Fe ( µg/L) 2290± 116 2266± 110 79± 127 <25 Dissolved Fe ( µg/L) 2312± 159 203± 133 <25 <25
*V l li d d d d i i f i l ll d b 3/10/04 d 8/26/04
Table 2. Results of Microibal Water Quality Analysis
Location HPC, CFU/mL AOB, MPN/mL Raw 4.1E+02 7.9E-01 Plant Effluent <10 <0.09 Media 4.5E+04 2.9E+03 Back Wash 6.6E+04 2.7E+02 Point 1 Well 4, 6, 8 2.2E+05 9.2E-01 Point 2 Well 4, 6, 8 3.2E+05 1.7E+01
60
Table 3. GenBank Relatives to 16S Clone Library Sequences Closest relative in GenBank (accession no.) Similarity (%)b % of clones in clone libraryc #OTUsd
Nitrospira sp. (AF035813) 96-98 29.6 2 Nitrospira sp. clone g6 (AJ224039) 95-99 20.8 2 Candidatus Nitrospira defluvii (EU559167) 99 0.5 1
Rhodospirillaceae bacterium LM22 (FJ455532) 94-98 7.2 5 Rhodospirillaceae bacterium L34 (FJ459988) 92 0.8 1 Sphingomonas sp.UF010 (AB426571) 99 2.1 1 Sphingomonas sp.EZ41 (EU591707) 95-97 2.7 4 Sphingomonas sp. MTR-71 (DQ898300) 95 1.1 1 Sphingomonadaceae bacterium HINF002 (AB426560) 95 0.5 1 Sphingomonas sp. HTCC503 (AY584572) 97 0.3 1 Sphingomonas sp. BAC151 (EU131005) 97 0.8 2 Hyphomicrobium vulgare (Y14302) 97 2.1 1 Hyphomicrobium sp. Ellin112 (AF408954) 95 0.5 2 Hypnomicrobium sp. KC-IT-W2 (FJ711209) 95 0.3 1 Nordella sp. P-63 (AM411927) 92 0.5 1 Bradyrhizobium sp. KC-EP-S3 (FJ711219) 99 0.3 1 Other Alphaproteobacteria (FJ203515, AY945895, 87-99 5.1 7 AF236002, AJ630204, NR 026337, and AF498710)
Nitrosomonas sp. Nm86 (AY123798) 97-98 5.1 2 Nitrosomonas sp. Nm84 (AY123797) 96 4.3 1 Nitrosomonas sp. Is32 (AJ621027) 97 0.3 1 Candidatus Nitrotoga sp. HAM-1 (FJ263061) 99 4.8 1 Methylophilus sp. ECd5 (AY436794) 96-97 0.5 2 Bacterium TG141 (AB308367) 93-96 1.3 3 Other Betaproteobacteria (AJ252690, AB271046, 97-98 1.6 4 AM412133, and DQ386262)
Gammaproteobacteria (AJ233898) 93 0.5 1
Deltaproteobacteria (DQ295890, DQ145534, AB246770, 82-96 2.1 7 CP001359, AB245340, NR 025348, and NR 024781)
Flavobacterium ferrugineum (AM230484) 94 1.9 1 Other Bacterodietes (EF612324, AF070444, and GQ144415) 86-90 0.8 3
Pirellula sp. (X81942) 92 0.8 1 Other Plantomycetes (AY162118, and CU925984) 87-98 0.5 2
Uncultured Chloroflexi bacterium (FM253645) 98 0.3 1
Opitutusa sp. VeSm13 (X99392) 97 0.3 1
a classified as a Verrucomicrobacter
61 b Percentage of similarity between the cloned 16S gene and its closest relative in the NCBI database c 375 total clones in the clone library d 65 total OTUs
A
B
C
62 Figure 1. Scanning electron micrograph of biofilms on the surface of Greene County filter media. (a) Backscatter image and EDS spectrograph showing patchy coating protecting biofilms (b) exposed biofilm and associated extracellular matrix (c) unique spirochete morphology.
99 OTU8 (1) 60 OTU9 (40)
62 OTU7 (1)
OTU5 (3) 79 100 OTU6 (8)
99 OTU3 (1)
OTU4 (3)
93 99 Nitrosomonas oligotropha AF272406 Nitrosomonas nitrosa AJ238495
Nitrosomonas communis AF272399
95 Nitrosomonas halophila AF272398 Nitrosomonas europaea Z97861 65 99 OTU2 (2) 100 Nitrosomonas europaea AL954747
OTU1 (1)
84 Nitrosospira multiformis AF042171 99 Nitrosospira tenuis AJ298720
M.capsulatus str Bath AE017282
0.05 Figure 2. Neighbor-joining phylogenetic tree of amoA nucleic acid sequences of ammonia oxidizing bacteria. Bootstrap values are shown based on 2000 replicates with values below 50% not shown. Values in parentheses indicate number of sequences in the OTU. 5% evolutionary sequence distance is represented by the scale bar.
63
a
i
r
e
d
l
o
h
k
r u
U B n c la s s if ie a d r 2 A s f1 l le 1 p ila 6 h ph a ylo T th 7 Me
R ho do bia les
Nit ros om on ad ale s
Rhizobiales
s le a d a n o m U o nk g no in wn h Be p ta S
B ac te ro di et
es
N
i
t
r
o
s
p
i
r
a
e
P
l
a
n
t
o
m
y
c
e
t
e s
0.05
Figure 3. Neighbor-joining phylogenetic tree of 16S nucleic acid sequences extracted from full-scale backwash. Sub-trees were reduced to branches proportionate to the number of sequences. 5% evolutionary sequence distance is represented by the scale bar.
64 SECTION 4.2 MANUSCRIPT II:
Biological Ammonia and Arsenic Oxidation in a Pilot-scale Rapid Sand Filter for Drinking Water Treatment
Colin White1 and Darren A. Lytle*2
1University of Cincinnati, Department of Biological Sciences Cincinnati, OH 45221
2United States Environmental Protection Agency, ORD, NRMRL, WSWRD, TTEB, Cincinnati, Ohio 45268
Abstract
The removal of ammonia from source water entering a drinking water distribution system is desirable, as excess levels have been correlated with nitrification, chlorine demand, corrosion, and biological re-growth. Several technologies exist to remove ammonia with recent interest in biological processes. The objective of this study was to design, operate, and evaluate the oxidation of ammonia and arsenic in pilot-scale, biologically active, rapid sand filters. Two pilot filters were seeded with backwash from a full-scale biologically active filter and ammonia oxidation observed approximately 40 days later. The filters were loaded with, and fully oxidized, approximately 1.3 mg/L ammonia-nitrogen and 95 µg/L arsenic at 3.5 gpm/sqft (EBCT 7.6 min).
Backwash water had a direct effect on nitrification recovery. Heterotrophic plate counts were lower in filter effluents compared to the influent and total coliforms were not detected in the filter effluent or influent. DGGE analysis confirmed that the microbial community shifted 22 after arsenic addition. This technology proved to mitigate excessive source water ammonia both reliably and efficiently as well as serve as a chemical-free method of arsenic oxidation for subsequent arsenic removal via sorption.
65 Introduction
Free ammonia in source waters has been linked to several undesirable
characteristics in finished drinking water and drinking water distribution systems.
Particularly, distribution system corrosion, taste and odor issues, chlorine demand,
biological growth, and nitrification are of primary concern (Zhang, 2008; Lytle, 2007).
Biological nitrification is the oxidation of ammonia to nitrite and nitrate. The
United States Environmental Protection Agency (US EPA) has set a source water
maximum contaminant limit (MCL) for nitrite and nitrate at 1 µg/L and 10 µg/L,
respectively. These products of nitrification are toxic to humans and cause blue baby
syndrome. Water leaving a treatment plant must not contain levels of nitrate and nitrate
above these values; however, finished water in the distribution system is not required to
be monitored. As free ammonia is not required to be removed from source waters, it may enter the distribution system and serve as substrate for nitrification leading to levels of nitrite and nitrate above the MCL.
Arsenic is commonly found co-exist with ammonia in ground water (Nordstrom
2001; Berg 2001; Mandal 2002; Nickson, 2002). The presence of arsenic in drinking water has many negative health effects, including being a known carcinogen (Mass,
1992; Smith, 1992). Recent data have prompted the United States Environmental
Protection Agency (USEPA) to reduce the arsenic standard in drinking water from 50 mg/L to 10 mg/L. Numerous studies have shown that arsenic can effectively be removed from water using iron-based treatment processes (e.g., iron based adsorption media, iron coagulation and iron removal) (Joshi, 1996; Edwards, 1994; Sorg 1978). Arsenic is present in two oxidation states in water. As(III) or arsenite and As(V) or arsenate.
66 Arsenite is considered more mobile in the environment, toxic, and difficult to remove by water treatment processes due to the lack of charge (Edwards, 1994). Therefore, in ground water containing Arsenite, arsenic should be oxidized to arsenate to achieve optimal arsenic removal. Oxidants such as chlorine and potassium permanganate can rapidly oxidize arsenite to arsenate, however aeration (oxygen) is not an effective oxidant
(Kim, 2000; Na, 2007; Eary, 1990). As ammonia exerts a high chlorine demand on the system, the consumption of oxidant for subsequent arsenic oxidation leads to a decrease in arsenic removal efficiency (Ghurye, 2001). Both the removal of excess ammonia and arsenic from source water is beneficial in drinking water treatment. Biological processes exist, and are common, in wastewater treatment facilities for the mitigation of excess ammonia, but there are few reports in the United States of biological processes for drinking water treatment. Moreover, to the knowledge of the authors, biological arsenic oxidation has not been intentionally implemented as a pre-oxidation step for arsenic removal in drinking water treatment. Lytle et al. (2007a and 2007b) reported on a full- scale system capable of concurrently oxidizing arsenic and ammonia. The study concluded that biological processes were responsible for the observed oxidation.
Biological water-treatment processes have been utilized in much of Europe for decades, but face resistance in the U.S. due to the negative perception of bacteria and limited data on operational reliability (Bouwer, 1988). Physicochemical processes exist for the removal of ammonia, but biological processes are believed to an efficient and cost effective alternative. There are several advantages of biological treatment such as reduced biodegradable organic matter (BOM), reduced biological growth in the
67 distribution system, and lower chlorine demand (Rittmann, 1984). The reduction of these three factors leads to biologically stable drinking water.
The objective of this study was to design and operate two pilot-scale rapid sand filters and evaluate their ability to oxidize ammonia and arsenic. The filters were seeded using water obtained from the full-scale filter previously described by Lytle et al.
(2007a). The effect of arsenic loading and backwash on ammonia oxidation was examined. Culture- and non-culture-based microbiological methods were used to evaluate the microbial community response.
Materials and Methods
Pilot filter design
The pilot system was constructed from common plumbing parts found at a local hardware store. Two 2.5 inch by 72 inch type PVC pipes were used as the filters. Filter 1
(F1) was filled to 35 inches with sand-anthracite atop a 5 inch support bed composed of half to quarter inch diameter gravel. Filter 2 (F2) was filled with 35 inches of half- to quarter-inch gravel atop 5 inches of 1 inch diameter gravel. The filters were filled via gravity and were operated in down flow using a Cole-Parmer gear pump (model 75211-
42) and controller to maintain effluent flow.
The filters were fed with water via gravity utilizing a constant head tank (Figure
1). Dechlorinated municipal tap water was fed to the filters after in-line injection of sufficient ammonium chloride (Fisher #A661) and sodium arsenite (Mallinckrodt #7392) to give a final concentration of 1 mg/L and 100 ug/L, respectively.
Source water for seeding
68 Lytle et al (2006) documented that the seeding method and type of seeding water did not affect the timeframe to establish nitrification in a pilot-scale filter. The filters were therefore seeded with 3L of full-scale filter backwash over a two-day period. Half of the 3L volume was poured into each filter and pumped through the media bed to within one inch of the top of the media and allowed to sit overnight. The remaining seed water was stored at 4oC and used the next day in the same manner. After the two-day
seeding process the filters were made operational.
Pilot filter operation
Flow rates were measured weekly using a stopwatch and graduated cylinder.
Filters were backwashed after reaching greater than 30 inches of headloss. Influent water
was used for backwashing. Backwash was performed using a flow rate sufficient to
fluidize the bed to 50% expansion (~2 L/min) for 7 minutes. Light backwashes were
performed in the event of filter dryout or air binding. In the event of a dryout, the filter
was backwashed lightly (less than 50% expansion) for no longer than 5 minutes. This
proved to be sufficient to remove air pockets that may cause water channeling and air
binding. Filters were shut off on weekends and headloss measured daily. Chlorinated
backwash consisted of municipal tap water spiked with 1 mg/L of free chlorine. Arsenic
injection started 40 days after the onset of complete nitrification.
Analytical measurements
Chemical samples were collected from the common influent and effluent lines of
each filter. Every effort was made to collect daily chemical samples, but circumstances
existed such that this goal was not always attainable. The pH and dissolved oxygen (DO)
was measured with a Hach Company (Loveland, CO) EC40 bench-top pH/DO meter
69 (Model 48600). The instrument was standardized for pH and DO daily using a two-point
calibration with pH 7 and 10 standard solutions and saturated tap water, respectively
(Whatman, Hillsboro, OR). All metals were measured with a Thermo-Fisher (Waltham,
MA) 6000 purged inductively coupled argon plasma spectrometer (ICAP) following
method 200.7. Ammonia-nitrogen, nitrate-nitrogen, and nitrate-nitrogen were measured using a Westco SmartChem autoanalyzer following methods 350.1 and 353.2. Total organic carbon was measured using combustion analysis following standard method 5310
C. Total alkalinity as CaCO3 was measured using potentiometric titration following
standard method 2320.
HPLC/ICP-MS for arsenic speciation was completed with an Agilent 1100 liquid
chromatograph (Agilent Technologies, Palo Alto, California) equipped with binary
HPLC pump, an autosampler, a 100!L sample loop, and a vacuum de-gasser system.
Reversed phase chromatography was performed with a ZORBAX EclipseXDB-C18 column (5mm "4.6mm id"250mm) from Agilent Technologies (Santa Clara, CA). The
ICP-MS used was an Agilent 7500ce (Agilent Technologies, Toyoko, Japan).
A anion exchange column following the procedure published by Edwards et al.
1997 was used for routine arsenic speciation and quantification by ICP/MS following method 200.8.
Microbial analysis
Heterotrophic plate counts were performed in duplicate using 10 fold dilutions following method 9215C using R2A (Remel) media. Plates were incubated at 28 oC for 7 days then counted. Total coliforms were cultured following method 9223B.
70 Total nucleic acid was extracted from 100 ml of 0.2 µm filtered influent water and
1.2 g of full-scale and pilot-scale filter media collected during filter operation. All samples were processed using the Epicentre MasterPure extraction kit (Madison, WI) with beadbeating. Denaturing gradient gel electrophoresis (DGGE) fragments were amplified via PCR using previously described primers and cycling conditions using 20 ng of template per reaction (O’Sullivan, 2008). A 40-60% linear urea denaturant gradient was formed in a 6% polyacrylamide gel. Approximately 500 ng of PCR product was loaded onto the gel, electrophoresed for 12 hours at 60oC, and imaged using GelStar.
Scanning Electron Microscopy
A small amount of filter media was fixed in a pH 7.2 cacodylate-buffered 1% paraformaldehyde 2.5% glutaraldehyde mixture. The media was post-fixed in 1% OsO4 and dehydrated in an ethanol series followed by Hexamethyldisilazane (HMDS) for 1 hour, and then air dried in a desiccator. The media was coated with gold/palladium prior to scanning electron microscopy (SEM). Samples were viewed using a JEOL 6490LV
(Peabody, MA) SEM at 25 kV under high vacuum.
Results
Average water quality data is presented in Table 1 and filtration details in Table 2.
Influent ammonia averaged 1.28 mg/L. Ammonia was oxidized in both filters to nitrite with a 0.3 mg/L nitrogen balance discrepancy overall (Figure 2). Filter 1 and 2 consumed 3.90 and 4.03 mg/L of oxygen, respectively. DO consumption was below that of the 4.57 mg O2/mg NH4-N previously reported and stoichiometrically calculated (EPA
1975). Municipal tap water provided a sufficient source of influent alkalinity for the
71 nitrification reaction. The alkalinity fluctuated over the duration of the study without any
detrimental effect on nitrification (Figure 5). The alkalinity was low enough, not to
resist changes in pH due to hydrogen ion production during nitrification. The drop in pH
averaged 0.88 units for both filters.
Alkalinity consumption was approximately 7.8 mg/L of CaCO3 for both filters,
lower than the theoretical of 9.13 mg/L of CaCO3 based on 1.28 mg/L NH4-N. TOC was
not considerably consumed in the filters averaging only a 0.015 mg/L decease compared
to the influent (Table 1).
After initial seeding, ammonia levels dropped to below detection (<0.03 mg/L) after day 40 (data not shown) and averaged 93% oxidation for the duration of the study.
Perturbations in complete ammonia oxidation were seen after day 60, but ammonia
breakthrough was never observed (Figure 2). These perturbations could not be correlated
with backwashing or water chemistry. Temperature, pH, and DO on and around these
days were average.
Nitrite levels in both filters remained near the detection limit (<0.01 mg/L) for the
duration of the study (Figure 3). Influent nitrite levels increased after day 30 and
remained unstable to the end of the study spiking as high as 0.56 mg/L (Table 1).
Backwashing did not have an affect on ammonia or nitrite oxidation in either filter.
Nitrate levels in the influent did not increase over time (Figure 4). Influent nitrate
averaged 0.91 mg/L over the duration of the study with effluent nitrate averaging 2.18
mg/L (Table 1).
72 Both filters accumulated headloss over the duration of the study (data not
shown). Filter 2 averaged an accumulated headloss of 6 inches in 2 weeks. Headloss in
filter 1 headloss much more rapidly, requiring backwashing every 7-10 days.
Average effluent HPCs were lower than influent HPCs over the duration of the study (Table 1). Effluent HPC spikes were expected but not observed with time. Results
of coliform cultures are not presented, as these assays did not detect any organisms.
After the first backwash, nitrification recovery after backwash was monitored.
Both filters regained approximately 85% complete ammonia oxidation after 3 bed
volumes (EBCT ~10 minutes) (Figure 6). During backwash, filter plugs were observed
in the top 3 to 5 inches of media in filter 1.
Filter 1 was backwashed with chlorinated water to observe nitrification recovery
time. This experiment was carried out at a loading rate of 3.5 gpm/sqft (EBCT 8
minutes). Chlorinated backwash water increased the ammonia oxidation recovery time
from approximately 22 minutes to over 12 hours (Figures 6 and 7).
Both filters oxidized As (III). Figure 8 shows arsenic speciation in the influent
and effluent of both filters. Dimethylarsinic acid (DMA) and methylarsonate (MMA)
were not detected in the influent or effluent. Speciation of arsenic by anion exchange
showed that influent As (III) became oxidized in the filter effluents after 22 days of
arsenic addition. Complete arsenic (III) oxidation was not observed though oxidation did
average grater than 70% for both filters over the duration of the arsenic addition (Table 1,
Figure 8).
Electron micrographs show an extensive biofilm network on the surface of media removed from the pilot-filters (Figure 10). These biofilms appeared patchy over the
73 surface of the anthracite of filter 1, but more uniform on gravel from filter 2 (Figures 10b
and 10c).
Molecular microbial analysis of effluent, influent, and seed water show distinct population changes. DGGE of full-scale filter media used for seeding resolved several bands not present in the filter effluents or influent, as well as comparing between filters
(Figure 9a). UPGMA analysis of DGGE fingerprints show filter media samples
clustered based on the presence of arsenic in the influent water (Figure 9b) after only 22
days of arsenic addition. At this time, As (III) was nearly 90% oxidized to As (V) in
filter effluent. Prior to arsenic addition, filters 1 and 2 were 85% similar while after
arsenic addition, were 62% similar. Seed media was 30% similar to all filter media
samples. Influent water was 37% similar to each other. All samples were approximately
25% similar. Each band of the DGGE can be interpreted a one OTU, and as such, the number of OTUs decreased from 20 in the seed water to 14 and 12 in F1 and F2 prior to arsenic addition, respectively. After arsenic addition OTUs in F1 and F2 were 16 and 10, respectively. Influent water without and with arsenic contained 6 and 5 OTUs, respectively. DGGE band intensity increases can be correlated to organism density. No specific bands were common to media samples taken after arsenic addition nor did any
bands after arsenic addition match seed water bands.
Discussion
Occasional fluctuations in ammonia oxidation could not be directly linked to
changes in water chemistry and backwash frequency. Several studies of biological
nitrification have reported on unstable ammonia oxidation where complete nitrification
74 decouples inexplicably (Wagner, 2002; Grahm, 2007, Jonsson, 2000). Grahm et al
(2007) reason that a fragile multalistic relationship exists between nitrite oxidizers and ammonia oxidizers. This fragile relationship is the source of sudden ammonia and nitrite buildup in nitrifying systems. As nitrite oxidation fails, the buildup of toxic nitrite inhibits ammonia oxidizers leading to ammonia breakthrough in the filters. However,
this is not likely the explanation for the data observed here because there was no buildup
of nitrite prior to ammonia oxidation, though our sampling strategy may have missed it.
The observation that influent nitrite increased coupled with the fact that influent
ammonia did not decrease proportionally suggests that the increase in nitrite was do to
the presence of nitrite in the municipal feed water. This increase in nitrite was of no
consequence as both filters were able to oxidize the increased nitrite load, suggesting the
filters have greater potential for nitrite oxidation even after only being exposed to an
average of 1.28 mg/L NO2-N following the ammonia oxidation step.
Headloss buildup did not affect nitrification. Being biologically active, filter
ripening, and headloss are a key variable for determining a backwash schedule. Filter 2,
being composed of larger media than filter 1, did not build significant headloss over time.
This is an important design criterion, as utilities may desire to pilot various media sizes to
best suit their needs and minimize costs associated with backwashing. Unpublished EPA
pilot studies have shown larger media, up to 0.5 inch gravel, is sufficient for biological
activity and nitrification at loading rates of 2 gpm/sq ft. This is of practical importance as
larger gravel can reduce clogging in waters with high iron. Thus, it may be beneficial to
implement such technology as a biological contactor, rather than filter, prior to filtration
to prevent downstream biofouling reduce filter backwashing.
75 Backwashing filter 1 with chlorinated water increased the ammonia oxidation recovery from 22 minutes to 13 hours (>85% oxidation). This is an important observation for utilities implementing biological treatment as a change in infrastructure may be required to store non-chlorinated water for backwashing. As the backwash frequency was low due to slow headloss buildup, it is possible that in waters with high organic carbon and iron a more aggressive backwash schedule may be required, though some reports suggest that frequent backwashing keeps the biofilm thin, allowing increased diffusion of water to the organisms. Along the same lines, coupling the observation of filter plugs in the top 3-5 inches of media (filter 1) and previous reports incorporating filter depth profiling, higher loading rates may be obtainable. First, as the loading rate increases the microbial population in the filter plug will be pushed lower into the filter, eventually colonizing a greater portion of the filter and increasing the reaction surface area (Andersson, 2001). Second, the diffusion of reactants to the microorganisms growing in biofilms on the surface of the media, increase due to the thinning of the biofilm as a result of increased hydraulic shear (Mohle, 2007). Whichever may occur, the filters are believed to have a greater capacity for microbial activity.
The capability of the filters to oxidize As (III) was examined. Previous research has shown that arsenic can be oxidized from As (III) to As (V) by diverse bacteria
(Inskeep 2007; Oremland, 2003; Macur, 2004). Here, from the start of the arsenic feed and without the use of a strong chemical oxidant, approximately 70% and 80% of influent arsenic was oxidized to As (V) in filters 1 and 2, respectively. This capacity to co-oxidize As (III) coupled with the many conceivable co-metabolic pathways present in a biologically active filter, may add to the operational flexibility and applicability in
76 drinking water treatment. Whether the As (V) would be removed in the filter with Fe
(III) or require a subsequent step needs to be studied, though single filter biological As
(III) oxidation and Fe (III) mediated As (V) removal has been reported (Lytle , 2007).
The results of the DGGE profiles show a decrease in OTUs between full-scale
and pilot-scale filters. This result, coupled with water quality analysis, suggests that the synthetic water used for piloting lacked many of the constituents present in the raw ground water feeding the full-scale plant. Thus, organisms present on the full-scale plant media used for seeding, may have been selected against and could not persist in such an
oligotrophic environment. This is also supported by the fact that the influent water
(dechlorinated municipal tap water) is biologically filtered and any organic carbon
present can be assumed to be non-bioavailable, as it has previously been degraded by
microbes. Though the numbers of OTUs per filter differed, both filters performed well,
suggesting that the different media profiles of the filters had no negative impact on
ammonia or arsenic oxidation. Another factor for the drop in OTUs observed between
the full-scale and pilot-scale filters is the difference in loading rate. It is possible that the
higher loading rate experienced in the pilot-system (1.5 gpm/sqft greater) may have
eliminated organisms not capable of forming strong biofilms. This could also explain
inter-filter OTUs differences as the finer media of filter 1 may provide denser biofilm
better protected from shear forces associated with backwashing.
The fact that no OTUs were common between the seed water and media after
arsenic addition, suggests that organisms other than those from the seed water may be
oxidizing arsenic in the pilot filters. As arsenic oxidizing bacteria are diverse, perhaps
77 influent microorganisms, capable of arsenic oxidation, out competed the seeded arsenic
oxidizers in the much more oligotrophic conditions of the pilot filters.
Though genus identification is possible through sequencing of excised DGGE
band products, we plan to study theses samples in greater detail using clone library
analysis and/or pyrosequencing techniques.
The lack of detection of indicator Coliforms suggests the filters do not harbor or
are a source of pathogens, though this may depend on unique source waters. Together
these results with prior reported data show that biological processes for drinking water
treatment are practical, efficient and safe.
The effect of influent water quality is an important variable that needs to be
studied in further detail. For instance, a nutrient rich water may support the growth of
other organisms and change how the filters operate. Future studies will address the
impact of loading rate, backwash frequency, and organic carbon load on the microbial
structure and operational functionality of the filters.
Acknowledgments
The authors would like to acknowledge fellow USEPA staff, Daniel Williams for
pilot assistance, Keith Kelty, Brittany Almassalkhi, Amy Reed, Andrea Galloway, and
Bill Kayler for analytical support; Laura Boczek, Cliff Johnson, and Jill Hoelle for
microbial analysis support; and John Sorrell for DGGE assistance. Any opinions expressed in this paper are those of the authors and do not necessarily reflect the official position and policies of the U.S. EPA. Any mention of products or trade names does not constitute recommendation for use by the U.S. EPA
78 References
Berg M, Tran HC, Nguyen TC, Pham HV, Schertenleib R, Giger W. (2001) Arsenic contamination of groundwater and drinking water in Vietnam: a human health threat. Environ Sci Technol 35:13 2621-6.
Bouwer EJ and Crowe PB (1988) Biological Processes in Drinking Water Treatment. Jour AWWA 80:82 93.
Eary LE and Schramke JA (1990) Rates of Inorganic Oxidation Reactions Involving Dissolved Oxygen. Chem Mod Aqs Sys II. 379-396.
Edwards M (1994) Chemistry of arsenic removal during coagulation and Fe-Mn oxidation Jour. AWWA 86:9.
Edwards M, Patel L, McNeill, HW (1997) Consideration in arsenic analysis and speciation. Jour AWWA 90:3 103-113.
Joshi A. and Chaudhuri M (1996) Removal of Arsenic from Groundwater by Iron- Oxide-Coated Sand. ASCE J. Environmental Engineering 122:8 769-771.
Ghurye, G. and Clifford D. (2001) Laboratory Study On The Oxidation Of Arsenic III to Arsenic V. U.S. Environmental Protection Agency, Washington, D.C., EPA/600/R- 01/021 (NTIS PB2001-108523).
Graham DW, Knapp CW, Van Vleck ES, Bloor, K, Lane TB, and Graham CE (2007) Experimental demonstration of chaotic instability in biological nitrification. ISME Journal 1 385-393.
Inskeep WP, Macur RE, Hamamura N, Warelow TP, Ward SA, Santini JM (2007) Detection, diversity and expression of aerobic bacterial arsenite oxidase genes. Environmental Microbiology 9:4 934-943.
Jonsson K, Grunditz C, Dalhammar G, Jansen JL (2000) Occurrence of nitrificaion inhibition in Swedish municipal wastewater. Water Res 34 2455-2462.
Kartinen EO and Martin CJ (1995) An overview of arsenic removal processes, Desalination 103:1-2 Proceedings of the American Desalting Association 1994 Biennial Conference and Exposition Membrane and Desalting Technologies, November 1995, Pages 79-88.
Kim MJ and Nriagu JO (2000). Oxidation of arsenite in groundwater using ozone and oxygen. Sci Total Environ 247 71-78.
Lytle DA, Sorg SJ, Muhlen C, Wang L, Rahrig M and French K (2007) Biological nitrification in a full-scale and pilot-scale iron removal drinking water treatment plant
79 J Water SRT-AQUA 56:2 125.
Lytle DA, Chen AS, Sorg TJ, Phillips S, French K (2007) Biological As(III) Oxidation in Water Treatment Plant Filters. Jour AWWA 99:12 72-86.
Lytle DA, Sorg TJ, Snoeyink VL (2005) Optimizing Arsenic Removal During Iron Removal: Theoretical and Practical Considerations. J Water SRT-Aqua 54 545-560.
Macur RE, Jackson CR, Botero LM, Mcdermott TR, Inskeep WP (2004) Bacterial Populations Associated with the Oxidation and Reduction of Arsenic in an Unsaturated Soil. Environ Sci Technol 38:1 104-111.
Mandal BK and Suzuki KT. (2002) Arsenic round the world: a review. Talanta 58:1 201- 35.
Mass MJ (1992) Human carcinogenesis by arsenic. Env Geochem and Health 14:2 49-54.
Mohle RB, Langemann T, Haesner M, Augustin W, Scholl S, Neu TR, Hempel DC, Horn, H (2007) Structure and shear strength of microbial biofilms as determined with confocal laser scanning microscopy and fluid dynamic gauging using a novel rotating disc biofilm reactor. Biotech Bioeng 98:4 747-755.
Na L, Fan M, Van Leeuwen J, Saha B, YANG H, HUANG CP (2007) Oxidation of As(III) by potassium permanganate. J Env Sci 19:7 783-786.
Nickson, R.T, J.M McArthur, P. Ravenscroft, W.G. Burgess, K.M. Ahmed. Mechanism of Arsenic Release to Groundwater, Bangladesh and West Bengal. Applied Geochemistry 15 (2002) 403-413.
Nordstrom, DK (2002) Worldwide Occurrences of Arsenic in Ground Water. Science 296:5576 2143-2145.
O'Sullivan LA, Webster G, Fry JC, Parkes RJ, Weightman AJ (2008) Modified linker- PCR primers facilitate complete sequencing of DGGE DNA fragments. J Microb Methods 75:3 579-58.
Oremland RS, Hoeft SE, Santini JM, Bano N, Hollibaugh RA, Hollibaugh JT (2002) Anaerobic Oxidation of Arsenite in Mono Lake Water and by a Facultative, Arsenite Oxidizing Chemoautotroph, Strain MLHE-1. Appl. Environ. Microbiol 68 4795.
Rittmann BE and Snoeyink VL (1984) Achieving Biologically Stable Drinking Water. Jour AWWA 76:10 106-114.
Smith AH, Hopenhayn-Rich C, Bates MN, Goeden HM, Hertz-Picciotto I, Duggan HM, Wood R, Kosnett MJ, Smith MT. (1992) Cancer risks from arsenic in drinking water. Environ Health Perspect 97:259-67.
80
Sorg, TJ and Logsdon. GS (1978) Treatment Technology to Meet the Interim Primary Drinking Water Regulations for Inorganics: Part 2. Jour. AWWA 70:7 379-393.
Wagner M and Loy A (2002) Bacterial community composition and function in sewage treatment systems. Curr Opin Biotech 13:3 2002 218-227.
Zhang Y, Griffin A, Edwards M (2008) Nitrification in premise plumbing: role of phosphate, pH and pipe corrosion. Environ Sci Technol 42:12 4280-4.
81
Figure 1. Schematic of pilot setup Arrows in lines indicate direction of water flow, open triangles indicate water line. (Not to scale)
82
Figure 2. Ammonia nitrogen concentrations from pilot filters. Each point represents one sample.
83
Figure 3. Nitrite nitrogen concentrations from pilot filters. Each point represents one sample.
84
Figure 4. Nitrate nitrogen concentrations from pilot filters. Each point represents one sample.
85
Figure 5. Alkalinity concentrations from pilot filters. Each point represents one sample.
86
Figure 6. Nitrification recovery after backwash. Each point represents one sample.
87
Figure 7. Nitrification recovery after chlorinated backwash.
88 1600
AsIII Influent 1400 F1 F2 1200 AsV 1000
800 C
600
400
200
0 0 100 200 300 400 Time (seconds)
Figure 8. Arsenic speciation of filter effluent water. (A) HPLC/ICP-MS chromatograph
of arsenic species. (B) Filter 1 effluent arsenic speciation. (C) Filter 2 effluent arsenic
speciation. (D) Influent Arsenic speciation.
89 M 1 2 3 4 5 6 7 8 9 10 11 12 13 14 A
B
Figure 9. DGGE analysis of microbial communities
(A) DGGE fingerprints. Lanes: Lane M, Molecular standard; Lane 1-2 Full-scale backwash media; Lanes 3-4 Filter 1; Lanes 5-6 Filter 2; Lanes 7-8 Influent; Lanes 9-10
90 Filter 1 after arsenic addition; Lanes 11-12 Filter 2 after arsenic addition; Lanes 13-14
Influent after arsenic addition. (B) UPGMA analysis of DGGE fingerprints (lanes designations same as above).
A B
C D
Figure 10. Scanning electron micrographs of media surface prior to arsenic addition (ET
13 days). (A) Biofilm on the surface of anthracite removed from filter 1 (B) High magnification of biofilm on the surface of anthracite removed from filter 1 (C) Biofilm on the surface of gravel removed from filter 2 (D) High magnification of biofilm on the surface of gravel removed from filter 2.
91
Table 1. Average water quality analysis data of pilot filters. Analyte Influent Effluent Filter 1 Effluent Filter 2 pH 8.00±0.26(57) 7.11±0.18(57) 7.13±0.19(57) Alkalinity (mg/L as CaCO3) 59.65±9.14(72) 51.66±8.79(72) 52.18±8.79(72) DO (mg/L) 9.02±0.44(58) 4.06±0.68(58) 4.21±0.77(58) Total Arsenic (µg/L) 85.9±0.03(50) 88.77±0.03(50) 88.08±0.03(50) Ammonia (mg/L as N) 1.28±0.15(69) 0.05±0.04(69) 0.08±0.13(69) Nitrite (mg/L as N) 0.14±0.12(64) 0.01±0(63) 0.02±0.01(64) Nitrate (mg/L as N) 0.91±0.32(64) 2.19±0.46(63) 2.16±0.53(64) TOC (mg/L) 1.07±0.19(24) 1.06±0.21(24) 1.05±0.16(24) HPC (cfu/ml) 41718±58741(11) 31389±23519(11) 34915±16202(11) Temperature (oC) 24.5±0.71(16255) 24.0±0.67(16255) 24.1±0.66(16255) *HPC values are geometric means Values are ± 1SD
Table 2. Filtration conditions Flow Rate Loading Rate EBCT Media (ml/min) (gpm/sqft) (min) Filter 1 Sand/anthracite 410 3.5 7.8 Filter 2 1/4" gravel 410 3.5 7.3
92 5. DISCUSSION
This study demonstrated that biological nitrification in a full-scale drinking water treatment filter was responsible for ammonia oxidation and that nitrification in pilot-scale filters was easily implemented and reliable. These results confirmed those of prior studies by Lytle and co-workers at the same treatment plant. Benefits for the use of biological processes for ammonia oxidation as an alternative to breakpoint chlorination are: lowered operational costs, increased biological stability in the finished water, and minimization of hazards associated with handling, transport, and use of chlorine. Lowered overall cost stems from the decrease in need to breakpoint chlorinate to achieve a free chlorine residual and decreased infrastructure maintenance due to biologically stable water. Bouwer and Crowe (1988) reported favorable opinions of biological treatment in European water districts due to lower costs. Achieving biologically stable drinking water is a key advantage of biological treatment. The reduction in biodegradable organic carbon (BOC) may decrease biological regrowth in the distribution system if these substances are consumed ahead of treatment. Biological regrowth may lead to reduced hydraulic capacity, increased corrosion, and pathogen growth and taste and odor issues in the distribution system (Rittmann, 1989). Lowering the chlorine demand of source water can also have a direct impact on the levels of halogenated disinfection by-products that have been shown to be carcinogenic. With excess chlorine in the finished water, taste and odor issues may also arise with the consumer. With the identification of xenobiotic degraders in the full-scale filters, the potential exists for degradation of complex organic materials. Though typically found in surface waters, such as rivers, the removal of pharmaceuticals and petroleum compounds may be enhanced by incorporating biological process ahead of UV or GAC processes. Pilot studies demonstrated ammonia was oxidized consistently without considerable negative effect from changes water chemistry or backwash. These pilot studies lend data to
93 the operational history and reliability of biological treatment. Though this work examined the molecular ecology of a full-scale system and the operation of a pilot-scale system, more experiments should be performed to study pathogen growth, detachment, kinetics, and reliability in varying source waters. Future experiments are described in the next chapter.
6. FUTURE WORK
As pilot-scale work is time consuming and sometimes complicated without online/real-time analysis, I purpose the following the following experiments:
With the dependence of nitrification on temperature and pH, experiments should be conducted using characteristic pHs of groundwater. This would be easily tested with the current pilot system as it contains a third feed port to which an acid or base may be fed. Heating or cooling the influent water or column beds may test the effects of temperature on nitrification. This will have two effects, by heating the bed, the impact on the microorganisms may be observed, and by heating the water, both the impact on the organisms as well as the impact on the ionization state of ammonia may be observed, though the latter harder to interpret.
With ammonia contamination in groundwater primarily resulting from fertilizer runoff, pesticide runoff is another concern. The injection of fertilizers concurrent with arsenic and ammonia would show the range of functionality of biological filtration.
Additionally, the application of TOC and other nutrients provide data to characterize the
94 operational constraints and help define influent water quality requirements for optimal
operation.
A more detailed study of the bacterial community shifts in response to changes in
operational parameters would help explain changes in filter performance and direct
engineers to remediation. The author planned many of these studies, but time limited
their design and execution.
7 REFERENCES
Ahn Y (2006) Sustainable nitrogen elimination biotechnologies: A review. Process Biochemistry 41:8 1709-1721.
Andersson A, Laurent P, Kihn A, Prevost M, Servais P (2001) Impact of temperature on nitrification in biological activated carbon (BAC) filters used for drinking water treatment. Water Research 35:12 2923-2934.
Anthonisen AC, Loehr C, PrakasamTBS, Srinath EG (1976) Inhibition of Nitrification by Ammonia and Nitrous Acid. Water Environment Federation. 48:5 835.
Awadalla FT, Striez C; Lamb K (1994) Removal of Ammonium and Nitrate Ions from Mine Effluents by Membrane Technology. Separation Science and Technology 29:4.
Bedard C and Knowles R (1989) Physiology, biochemistry, and specific inhibitors of CH4, NH4+, and CO oxidation by methanotrophs and nitrifiers. Microbiological Reviews 53:1 68-84.
Belser LW (1984) Bicarbonate Uptake by Nitrifiers: Effects of Growth Rate, pH, Substrate Concentration, and Metabolic Inhibitors. Appl. Environ. Microbiol. 48:1100- 1104.
Campbell CL, Dawes RK, Deolakkar S, Merritt MC (1958) Effects of certain chemicals in water on the flavor of brewed coffee. Food Research 23 575-79.
Cotruvo, JA (1981) Trihalomethanes (THMs) in drinking water. Environ. Sci. Technol. 15:3 268–274.
95 Demir A, Gunay A, Debik E (2002) Ammonium removal from aqueous solution by ion exchange using packed bed natural zeolite. Water SA. 28:3 329-335.
Du Q, Liu S, Cao Z, Wang Y (2005) Ammonia removal from aqueous solution using natural Chinese clinoptilolite. Separation and Purification Technology. 44:3 229-234.
[EC] European Commission. (1998b) Directive 98/83/EC. OJ L 330 33–54.
Family physician. A manual of domestic medicine, by physicians and surgeons of the principal London hospitals. New and enlarged edition ed. London, Paris, New York & Melbourne: Cassell and Company LTD, 1886.
Fleischacker SJ, Randtke SJ (1983) Formation of Organic Chlorine in Public Water Supplies. Jour AWWA 75:3 132-138.
Francis CA, Roberts KJ, Beman JM, Santoro AE, Oakley BB. (2005) Ubiquity and diversity of ammonia-oxidizing archaea in water columns and sediments of the ocean. Proc Natl Acad Sci U S A. 102:41 14683-8.
Geneva. World Health Organization. Ammonia. Environmental Health Criteria No. 54, 1986.
Hagopian DS and Riley JG (1998) A closer look at the bacteriology of nitrification. Aquacultural Engineering 18:4 223-244.
Hollocher TC, Kumar S, Nicholas DJD (1982). Respiration-dependent proton translocation in Nitrosomonas europaea and its apparent absence in Nitrobacter agilis during inorganic oxidations. J Bacteriol. 149 1013–1020.
Hussain Allem HMI and Sewell DL (1981) Mechanism of nitrite oxidation and oxidoreductase systems in Nitrobacter agilis. Current microbiology 5:5 267-272.
(INRS) Institut National de Recherche et de Securite de France. (1987) Ammoniac et solutions aqueuses, fiche toxicologique 16. Cahiers de notes documentaires 128 461-65.
Isirimah NO, Keeney DR, Dettmann EH (1976) Nitrogen Cycling in Lake Wingra1 J Environ. Qual. 5 182-188.
Jafvert CT, Valentine RL (1992) Reaction scheme for the chlorination of ammoniacal water. Environ. Sci. Technol. 26:3 577–586.
Kihn A, Andersson A, Laurent P, Servais P, Prévost M (2002) Impact of filtration material on nitrification in biological filters used in drinking water production. J Water SRT Aqua 51 35-46.
Knowles R (1982) Denitrification. Microbiological Reviews 46:1 43-70.
96
Koyuncu I, Topacik D, Turan M, Celik MS, Sarikaya HZ (2001) Application of the membrane technology to control ammonia in surface water. Water Sci. Tech. SA. 1:1 117-124.
Kumar S and Nicholas JD (1983) Proton Electrochemical Gradients in Washed Cells of Nitrosomonas europaea and Nitrobacter agilis. J Bacteriol. 154:1 65–71.
Kurama H, Poetzschke J, Haseneder R (2002) The application of membrane filtration for the removal of ammonium ions from potable water. Water Research 36:11 2905-2909.
Lin JT and Stewart V (1998) Nitrate assimilation by bacteria. Adv Microb Physiol. 39:1-30, 379.
Lytle DA, Sorg SJ, Muhlen C, Wang L, Rahrig M and French K (2007) Biological nitrification in a full-scale and pilot-scale iron removal drinking water treatment plant J Water SRT-AQUA 56:2 125.
Moore EM, (1951) Water Sewage Works 98:3.
Norddahl B, Horn VG, Larsson M, du Preez JH, Christensen K (2006) A membrane contactor for ammonia stripping, pilot scale experience and modeling. Desalination 199:1-3 172-174.
Papen H, von Berg R, Hinkel I, Thoene B, Rennenberg H (1989) Heterotrophic nitrification by Alcaligenes faecalis: NO2-, NO3-, N2O, and NO production in exponentially growing cultures. Appl. Environ. Microbiol. 55:8 2068-2072.
Pressley TA, Bishop DF, Roan, SG (1972) Ammonia-nitrogen removal by breakpoint chlorination Environ. Sci. Technol. 6:7 622–628.
Rimon A and Shilo M (1982) Factors which affect the intensification of fish breeding in Israel. Bamidgeh, 34(3): 87-100.
Rittmann BE, Huck PM, Bouwer C (1989) Biological treatment of public water supplies. Critical Reviews in Environmental Science and Technology 19:2 119 – 184.
Rittmann BE and Snoeyink VL (1984) Achieving Biologically Stable Drinking Water. Jour AWWA 76:10 106-114.
Siegrist H and Gujer W (1987) Demonstration of mass transfer and pH effects in a nitrifying biofilm. Water Research 21:12 1481-1487.
Simon J (2002) Enzymology and bioenergetics of respiratory nitrite ammonification. FEMS Microbiol Rev. 26:3 285-309.
97 Smith CJ, Nedwell DB, Dong LF, Osborn AM (2007) Diversity and abundance of nitrate reductase genes (narG and napA), nitrite reductase genes (nirS and nrfA), and their transcripts in estuarine sediments. Appl Environ Microbiol. 73:11 3612-22.
Smyth HF, Seaton J, Fischer L (1941) "The single dose toxicity of some glycols and derivatives." J. Ind. Hyg. Toxicol. 23 259-68.
U.S.EPA (1975) Process Design Manual for Nitrogen Control, Office of Technology, Washington D.C.
U.S. DHHS (2004) Department of Health and Human Services. Agency for Toxic Substances and Disease Registry. Public Health Statement. Ammonia ed. 2004.
U.S. EPA (1989) Environmental Protection Agency. Summary review of health effects associated with ammonia. Washington, DC: (EPA/600/8-89/052F), 1989.
Verhagen FJM, Laanbroek, HJ (1991) Competition for Ammonium between Nitrifying and Heterotrophic Bacteria in Dual Energy-Limited Chemostats. Appl. Environ. Microbiol. 57:11 3255-3263.
Ward BB (1987) Kinetic studies on ammonia and methane oxidation by Nitrosococcus oceanus. Archives of Microbiology 147:2.
Waton SW and Waterbury JB (1971) Characteristics of two marine nitrite oxidizing bacteria, Nitrospina gracilis nov. gen. nov. sp. and Nitrococcus mobilis nov. gen. nov. sp. Archives of Microbiology 77:3
Westerhoff P, Chao, Mash H, (2004) Reactivity of natural organic matter with aqueous chlorine and bromine, Water Research 38:6 1502-1513.
Witzel and Overbeck (1979) Heterotrophic nitrification by Arthrobacter sp. (strain 9006) as influenced by different cultural conditions, growth state and acetate metabolism. Archives of Microbiology 122:2.
Yamanaka and Sakano (2007) Oxidation of hydroxylamine to nitrite catalyzed by hydroxylamine oxidoreductase purified from Nitrosomonas europaea current microbiology 4:4 239-244.
Zhang Y, Triantafyllidou, S, Edwards, M (2008) Effect of Nitrification and GAC Filtration on Copper and Lead Leaching in Home Plumbing Systems. J. Envir. Engrg. 134:7 521-530.
Zhang Y, Griffin A, Rahman M, Camper A, Baribeau H, Edwards M. (2009) Lead contamination of potable water due to nitrification. Environ Sci Technol. 15:43(6):1890- 5.
98
APPENDIX
Quality Assurance and Quality Control*
*Adapted from the Quality Assurance Project Plan written by Colin White for the US EPA.
Quality control checks for Molecular techniques
Data collected as a result of this work will be used to assess microbiological communities in drinking water biofilms to assess potential health risks and disinfection strategies. The results of this work are intended to provide information on the: (1) microorganism relationships in biofilms and (2) assess integrate molecular techniques with standard microbiological techniques and drinking water quality data.
Molecular experiments and data collection will be performed under standard procedures and conditions for molecular biology in accordance with Current Protocols in Molecular Biology and manufacturer product guidelines. Standard microbiological techniques and data collection will be performed using standard procedures from Standard Methods for the Examination of Water and Wastewater
Because PCR methods will be used in the detection microorganisms in water, the QA/QC procedures outlined in the recent EPA publication (815-B-04-001) "Quality Assurance/Quality Control
Guidance for 'Laboratories Performing PCR Analyses on Environmental Samples" will be followed.
The objective of quality control in this project is to provide scientific data that will be repeatable.
PCR positivity
The accuracy goal for PCR is to detect microorganisms with high sensitivity and specificity. For sensitivity, the ultimate goal is to detect a low cell counts in water in the presence of interference materials without producing false negatives.
Specificity
For specificity, the goal is to detect only select microorganism DNA thus; a false positive is also not acceptable. Because only qualitative data will be generated, data from all PCR runs will be evaluated for the presence of false positive and false negative results, using positive and negative controls in each
99 PCR. PCR products will be sequenced to confirm result of the PCR diagnosis. Only data from PCR runs without false positive and false negative results will be acceptable.
DNA sequences
The accuracy goal for DNA sequencing is to generate DNA sequences free of errors. All PCR products from each sample will be sequenced in both the 5' and 3’ directions, and the sequences obtained will be compared with each other and those downloaded from the public sequence database GenBank. Any sequence differences will be verified through manual inspection of the electropherogram (track file from the automatic DNA sequencer), and if needed by sequencing a second PCR product from the same sample or DNA.
Major error sources and roadblocks
PCR contamination, poor PCR amplification, and PCR inhibition. The former results in false positives and is common to all PCR especially nested PCR `techniques. The latter two are more common to the analysis of environmental samples. Poor PCR amplification is usually a result of un-optimized PCR primers and conditions. There are many standardized approaches to address PCR contamination and poor
PCR amplifications which, will be used to minimize the occurrence of the two problems in this project we do not expect them to be major roadblocks in this project. What is more problematic and common to all
PCR methods for the analysis of environmental samples is the inherent richness of PCR inhibitors in water samples. The co-extraction of PCR inhibitors during DNA isolation can greatly reduce the sensitivity of
PCR detection, leading to false negative results. In this project, we have chosen to purify nucleic acids using beadbeating and ion-exchange chromatography. This method has been shown, and it can be expected to, produce DNA of high purity. We will also use a high concentration of non-acetylated bovine serum- albumin in PCR to neutralize residual inhibitors.
DNA extraction
100 DNA will be extracted from water samples using Epicentre Soil DNA extraction kits. This technique was shown earlier by us to effectively remove PCR inhibitors and has been used effectively in
PCR detection of microorganisms in various water samples and sediments.
DNA sequencing
To verify the results of genotyping, primer specificity, and to subtype specific organisms all positive products of the expected size and all PCR products of the unexpected sizes will be sequenced using an ABI 3100 genetic analyzer.
Work areas and work flow
This project is organized to maintain separate work areas for molecular techniques. These two areas are 1) a sample preparation/pre amplification area and 2) an amplification area. While the EPA
QA/QC PCR Guidance documents suggests that work involving PCR be divided into three separate areas, limitations on available space preclude the use of three areas during this project.
The sample preparation is where samples will be made ready for amplification (nucleic acid extraction) and reagent are added to the sample and QC samples, such as negative and positive controls.
A standard operating procedure will be instituted to ensure decontamination of work areas prior to each sample/control preparation as covered in Appendix A. These measures include the use of multiple nucleic acid destroying agests, such as DNA-Away, UV irradiation, 95% ethanol, and 10% bleach.
All samples and reagents will be stored in separate refrigerators/freezers and according to manufacturers’ guidelines. A dedicated set of pipettes will be used for sample preparation. Aerosol resistant tips will be used for all work.
DNA amplification work will flow in one direction with dedicated equipment and supplies used in each area. Gloves will be worn during all stages of the process.
Work will flow from the sample preparation/pre-amplification areas to the amplification area. Reagents and equipment maintained in the amplification area will not be moved into the sample preparation/pre- amplification area without undergoing decontamination procedures and testing to prevent contamination.
101 The amplification area will be physically located in a separate area form the sample preparation/pre-amplification area.
A Biorad Tetra2 PCR machine will be located in the amplification area.
Sample reactions tubes must never be open in the amplification area. Following amplification, all tubes will be disposed of in an autoclave and according to standard protocols for biohazard waste disposal.
Amplified tubes will be moved to the preparation/pre-amplification area following amplification. Any changes to the PCR setup will be noted in laboratory notebooks.
Positive Controls
The following positive controls will be used with each sample processing runs:
PCR positive control
To control for poor PCR amplification, one positive, control will be used in each PCR run. For
16S rDNA based PCR, the control we will use is DNA of E. coli. CN13, which is a laboratory strain microorganism. For species specific primers, a known DNA sample of the species of interest will serve as the positive control. These control DNA samples will be obtained from researchers at the US EPA.
Electrophoresis control
100-bp DNA size ladders will be used in each electrophoresis run to verify that the electrophoresis apparatus works properly and to verify correct PCR amplification size. This ladder will be purchased from
Alpha Innotech (Cat# 60-13069-00).
DNA sequencing control
The accuracy of DNA autosequencer is checked every thirty runs with DNA sequencing of the
PCR product of the 16S subunit rDNA gene of a control E. coli CN13 specimen. Although it is unlikely that the misincorporation of one or two nucleotides during PCR will lead to misdiagnosis, to reduce error rates, PCR products will be sequenced directly without cloning, because our experience indicates this practice has lower error rates than sequencing of cloned PCR products.
102
Negative control
The following negative controls will be used with each sample processing runs:
Method negative control
A method blank containing only reagent water and no microorganisms will be run with each sample processing and DNA extraction.
PCR negative control
To control contamination, 2 negative controls (no DNA template) will be used in each PCR run.
This will rule out the possibility of contaminated reagents.
Analytic replicates
To ensure accurate diagnosis, each sample will be analyzed by PCR at least two times, and the results will be confirmed by the sequencing of all positive PCR products from each sample. All sequencing will be done in both 5' and 3' directions, using sequencing primers for every 500 bp. If there is any discrepancy (sequences from forward or reverse sequencing, sequences from two PCR products do not match each other exactly) in the nucleotide sequence obtained from the 2 PCR products, a third PCR product from the sample will be sequenced to resolve the discrepancy. If only one positive PCR product from the two PCR replicates is available for sequencing and a unique sequence is obtained, DNA sequencing of more independent PCR products from the same sample will be conducted.
False positive/negative prevention
False positive prevention
The QA/QC procedures outlined in the recent EPA publication (815-B 04-001) "Quality
Assurance/Quality Control Guidance for Laboratories Performing PCR Analyses on Environmental
Samples" will be followed strictly to reduce the occurrence of false positive results. In addition to the use of sample, method or PCR negative controls to ensure there will be no introduced contamination, all
103 molecular analysis related to this project will be conducted in two designated rooms. Disposable gloves and white laboratory coat will be worn before sample handling. Sample preparation and DNA extraction will be done in areas separated from PCR preparation and DNA sequencing. Filter-plugged tips will be used in all pipetting of samples and reagents. Different pipettes will be used in sample preparation, PCR, gel electrophoresis, RFLP and sequencing. They will be decontaminated by treatment in Stratalinker before each use. Only disposable plastics free of nucleic acids and DNase and RNase will be used in DNA extraction and PCR. PCR preparation will be conducted in a laminar biological safety cabinet or PCR station, with sample loading occurring in a separated area.
False negative prevention
In addition to the use of positive controls, only analytical and preparational procedures published in the scientific community or standardized by us are used in this project. PCR methods used in the project have been optimized, and the PCR condition can be modified further if needed. The use of 400 ng/pl of non-acetylated bovine serum albumin in PCR will be employed to neutralize residual PCR inhibitors. If needed, DNA polymerase with know resistance to PCR inhibition, such as Tth, will replace Taq in PCR.
Testing, Inspection, and Maintenance Requirements
Biorad Tetrad2 Quality Control
The tetrad2 must have a unique tracking number. Laboratory notebooks and/or other project records must reference (by tracking number) the tetrad2 used in experiments.
The Tetrad2 must have a logbook associated with it, where maintenance and other relevant information is recorded. The logbook must track user names; user’s associated files, and programmed parameters for each of the user’s files.
Operation parameters will be reviewed prior to nay run to verify that the correct thermal cycle parameters are programmed and are being selected.
Sample blocks will be cleaned periodically. Individual sample wells will be cleaned with cotton swabs moistened with hydrogen peroxide. Cleaning must be recorded in the logbook.
104 Instrument calibration and frequency
Other instruments and equipment:
Other equipment involved in sample preparation/pre-amplification procedures includes but is not limited to, micropipetters, incubator, water bath, centrifuges, heat blocks, electrophoresis chambers and computers.
Micropipettes are subject to annual calibration/maintenance, usually by an outside service provider. Records of such service must be maintained by researchers. Individual pipettes must be traceable to the calibration records.
Temperature regulating equipment must be check on a routine basis based on branch or laboratory specific guidelines. Thermometers or thermocouples use to check temperatures must be traceable to a certified thermometer.
The accuracy of the DNA autosequencer is checked every thirty runs with DNA sequencing of the PCR product of small subunit rDNA gene of a control Escherichia coli specimen.
Graphite Furnace Atomic Absorption Spectrometry
The instrument will be calibrated daily and each time the instrument is set-up. The daily standard curves will consist of a blank and at least three standards within the linear dynamic range of the instrument.
Initial calibration verification must fall within the control limits of 90-110% of the known value. For continuing calibration verification, the calibration check sample and the calibration blank will be analyzed every ten samples and at the end of the run. An outside reference standard will also be run at least once per analysis run. The calibration blank must be less than the PQL. If the response for any analyte falls outside the control limits, a new standard curve will be prepared and all samples run since the last calibration verification will be reanalyzed. A record will be made of all calibration verification results.
Inductively Coupled Plasma Emission Spectrometry
Calibration procedures for the ICP will be similar to those outlined for the graphite furnace.
pH Measurements
105 pH meters will be calibrated daily using commercially purchased pH 7, pH 4, and pH 10 buffer solutions. If the calibration slope is not with 95 B 105%, the probe will be cleaned or replaced until the calibration is adequate.
Analytical Balances
Analytical balances will be calibrated on a routine basis with a set of certified weights and records will be kept in a logbook. The laboratory has service contracts on all balances that provide annual calibration by service personnel.
Kits and reagents
All purchased kits and reagents will be tracked in a laboratory notebook or on appropriate forms and maintained in project files.
Information tracked must include material name, supplier/manufacturer, lot number(s), date received or formulated, date opened, and expiration date.
Use of new materials must be noted in laboratory notebooks.
Quality control checks for microbiological techniques (cultures )and kinetics
Culturing technique and quality control will follow established EPA standard methods for the organism of interest (Table A1).
Kinetic experiment controls will consist of both sterile water dosed with ammonia/arsenite and sterilized concentrated water dosed with ammoinia/arsenite.
Inspection/Acceptance requirements for supplies and consumables
Nucleic acid extraction supplies and consumables will be inspected upon receipt by Mr. White.
Acceptance of these will be based upon visually determining that received material is consistent with project requirements and intact packaging. Items identified as damaged or contaminated will be declined.
Assay supplies and consumables will be inspected and accepted upon receipt in the laboratory if they are intact, sterile, DNase and RNase free, and if of enzymatic nature, that they were properly shipped
106 on ice. Expiration dates shall be marked on packages and logged appropriately. People responsible for proper receipt of materials are Mr. White.
Data Acquisition Requirements (Non-Direct Measurements)
Sequence data for project implementation will be obtained from public computer databases including:
DNA sequences from microorganisms will be obtained from Genbank (National Center for
Biotechnology Information) using BLAST software, supported by the National Institutes of Heath
and National Library of Medicine.
Literature will be obtained from the National Library of Medicine using Entrez PubMed
------
This section describes quality assurance/quality control (QA/QC) checks that will be used throughout the project.
Identified QA Objectives
Accuracy is broadly defined as how close the analyses will come to the true concentration in the sample.
The accuracy of measurements incorporating a standard reference material or a second source standard will be calculated as percent recovery.
% Recovery = 100% ! (Cs/Cmst) (5-1) where Cs is the measured concentration of the standard and Cmst is the actual concentration of the standard.
The accuracy of the analyses that use matrix spikes will be calculated by
% Recovery = 100% ! (Csp - Cmsa) / Cac (5-2) where Csp is the measured concentration of the spiked aliquot, Cmsa is the measured concentration of the sample and Cac is the actual concentration of the spiked aliquot. The accuracy of the samples that cannot be determined with Equations 5-1 and 5-2 will be calculated by the measurement bias.
Bias = Mb - Mk (5-3)
107 where Mb is the measurement with bias and Mk is the known value.
Precision is broadly defined as the scatter within any set of repeated measurements. For samples that are measured in duplicate, precision will be calculated as relative percent difference (RPD).
RPD = (C1-C2) / ((C1+C2) / 2) ! 100 (5-4) where C1 and C2 are the two measurements. For samples that are measured in triplicate or higher, the precision will be measured as the relative standard deviation (RSD).
RSD = (S / SM) ! 100 (5-5) where S is the standard deviation and SM is the sample mean. Precision of the measurements that cannot be calculated with 5-4 and 5-5 will be determined by absolute range (AR).
AR = |M1 - M2| (5-6) where M1 and M2 are the two measurements.
Completeness is broadly defined as what percent of samples are deemed to be valid (%C)
%C = (V / TOT) ! 100 (5-7) where V is the number of samples determined to be valid and TOT is the total number of measurements.
Table A1 shows the MDL, precision, accuracy and completeness for the ICAP, pH, and temperature analyses. If these criteria are met, the analysis will be accepte
Table A1. QA Objectives for Method Detection Limit, Precision, Accuracy and Completeness
Parameter Reporting
Units MDL Precision Accuracy Completeness
Inorganics
Alkalinity mg/L 1.0 2.4% 98-102% 95%
Aluminum mg/L 0.025 5.4% 98-102% 95%
Arsenic ug/L 2 3.7% 99-101% 95%
Calcium mg/L 0.010 2.9% 99-101% 95%
108 Conductivity µS/cm 1 5% 99-101% 95%
Copper mg/L 0.003 6.3% 99-101% 95%
Iron mg/L 0.003 3.0% 99-101% 95%
Manganese mg/L 0.035 3.6% 99-101% 95%
Manganese mg/L 0.0004 2.7% 99-101% 95%
Nickel mg/L 0.015 6.1% 99-101% 95%
Potassium mg/L 0.05 4% 97-103% 95%
pH pH units -- 0.1 -- 95%
Silicon mg/L 0.012 3.2% 97-103% 95%
Sodium mg/L 0.020 3.9% 99-101% 95%
Sulfur mg/L 0.030 9.6% 99-101% 95%
Zinc mg/L 0.001 5.4% 98-102% 95%
Organics
TOC mg/L 0.05 0.1 0.15 95%
Particulates
Turbidity NTU 0.04 0.02 95-105% 90%
Physical
Temperature oC -- 0.1oC 0.1oC 95%
Quality Control
Table A2 summarizes quality control checks for analytical measurements. The quality control
(QC) checks for ICAP, include duplicates, matrix spikes. Matrix spikes (MS) will be analyzed at a frequency of 1 per 20 samples. Acceptance criteria for the MS will be 85 to 155% recovery.
The QC checks for molecular methods will follow those specified in the QA/QC guidelines of the
EPA publication (815-B-04-001) “Quality Assurance/Quality Control Guidance for Laboratories
109 performing PCR Analyses on Environmental Samples” that include matrix spikes, positive controls, negative controls and laboratory blanks specified above, will be used to minimize errors.
Table A2. Summary of QC Checks
Acceptance Analysis QC checks Method Objective Frequency Criteria
Extraction Matrix spikes 1 per 20 ICAP Spiking efficiency and 85-155% recovery (MS) samples contamination
Laboratory 1 per Blank Qualitative PCR Blank & reaction Presence/absence Duplicate analysis Duplicates mixture
Laboratory Blank Quantitative 1 per Kinetics Blank & Presence/absence Duplicate analysis experiment Duplicates
Laboratory Blank Qualitative 1 per DNA Extraction Blank & Presence/absence Duplicate analysis extraction Duplicates
Laboratory Blank Qualitative 1 per HFUF Presence/absence Blank analysis filtration
Laboratory Blank Quantitative 1 per liter AOB Blank & Presence/absence Duplicate analysis of media Duplicates
Laboratory Quantitative 1 per E.coli Blank Presence/absence Blank analysis sample
Laboratory Quantitative 1 per Enterococci Blank Presence/absence Blank analysis sample
110 Laboratory Quantitative 1 per Coliforms Blank Presence/absence Blank analysis sample
Laboratory Aerobic Blank Quantitative 1 per Presence/absence Blank & Endospores Duplicate analysis sample % RPD <20% Duplicates
Laboratory Blank Quantitative 2 per Presence/absence HPC Blank & Duplicate analysis sample % RPD <20% Duplicates
Quantitative 1 per Presence/absence Electrophoresis MS Spiking and qualitative sample analysis
Matrix spikes
(MS), Blank Ammonia, Nitrite, Quantitative 1 per 10 Laboratory Duplicate 85-115% Spikes Nitrate analysis samples Blank & Spiking
Duplicates
111