anaerobic digestion for methane generation and ammonia … · anaerobic digestion for methane...

ww.sciencedirect.com

b i om a s s a n d b i o e n e r g y 5 6 ( 2 0 1 3 ) 4 9 3e5 0 5

Available online at w

http: / /www.elsevier .com/locate/biombioe

Anaerobic digestion for methane generation andammonia reforming for hydrogen production:A thermodynamic energy balance of a modelsystem to demonstrate net energy feasibility

David M. Babson a, Karen Bellman b, Shaurya Prakash b,*,Donna E. Fennell a,**aDepartment of Environmental Sciences, Rutgers University, 14 College Farm Road, New Brunswick, NJ 08901,

United StatesbDepartment of Mechanical and Aerospace Engineering, The Ohio State University, 201 W. 19th Ave., Columbus,

OH 43210, United States

a r t i c l e i n f o

Article history:

Received 6 October 2012

Received in revised form

6 May 2013

Accepted 24 May 2013

Available online

Keywords:

Anaerobic digestion

Ammonia

Bioenergy

Bioammonia

Hydrogen

Anaerobic digestion-bioammonia

to hydrogen (ADBH)

* Corresponding author. Tel.: þ1 614 688 404** Corresponding author. Tel.: þ1 848 932 574

E-mail addresses: [email protected] (S0961-9534/$ e see front matter ª 2013 Elsevhttp://dx.doi.org/10.1016/j.biombioe.2013.05.

a b s t r a c t

During anaerobic digestion, organic matter is converted to carbon dioxide and methane,

and organic nitrogen is converted to ammonia. Generally, ammonia is recycled as a fer-

tilizer or removed via nitrificationedenitrification in treatment systems; alternatively it

could be recovered and catalytically converted to hydrogen, thus supplying additional fuel.

To provide a basis for further investigation, a theoretical energy balance for a model sys-

tem that incorporates anaerobic digestion, ammonia separation and recovery, and con-

version of the ammonia to hydrogen is reported. The model Anaerobic Digestion-

Bioammonia to Hydrogen (ADBH) system energy demands including heating, pumping,

mixing, and ammonia reforming were subtracted from the total energy output from

methane and hydrogen to create an overall energy balance. The energy balance was

examined for the ADBH system operating with a fixed feedstock loading rate with C:N

ratios (gC/gN) ranging from 136 to 3 which imposed corresponding total ammonia nitrogen

(TAN) concentrations of 20e10,000 mg/L. Normalizing total energy potential to the

methane potential alone indicated that at a C:N ratio of 17, the energy output was greater

for the ADBH system than from anaerobic digestion generating only methane. Decreasing

the C:N ratio increased themethane content of the biogas comprising primarily methane to

>80% and increased the ammonia stripping energy demand. The system required 23e34%

of the total energy generated as parasitic losses with no energy integration, but when

internally produced heat and pressure differentials were recovered, parasitic losses were

reduced to between 8 and 17%.

ª 2013 Elsevier Ltd. All rights reserved.

5; fax: þ1 614 292 3163.8; fax: þ1 732 932 8644.. Prakash), [email protected] (D.E. Fennell).ier Ltd. All rights reserved.024


1. Introduction

1.1. Anaerobic digestion processes and applications

In recent years, major attitudinal shifts have occurred in

modern society to allow wastes to be considered resources

rather than just materials requiring disposal [1e3]. Waste as a

resource can be envisioned as a means of reducing energy

consumption by minimizing the need for producing raw ma-

terials, reducing pre-disposal processing, and extracting us-

able energy from the waste as a feedstock [4e6]. One

established technology for extracting energy from waste is

anaerobic digestion (for recent reviews see Refs. [7e11]).

Anaerobic digestion of crop biomass, agricultural wastes

and residuals, and source-separated mixed organic wastes

have been employed at full scale for decades in Europe (nearly

200 digesters, 2010 [12]), China (10,000 digesters, 1986 [13]) and

India (2,000,000 digesters, 2000 [14]), among other places [15,16].

These digesters operate to generate biogas, comprising pri-

marily of methane, as a fuel source. In the US, anaerobic

digestion is primarily used for wastewater treatment plant

(WWTP) sludges [17], animal manures [18], andmunicipal solid

waste (MSW) in landfills [19]. There aremore than 500 large (>5

million gallons per day) municipal wastewater treatment fa-

cilities [20] and 176 animalmanure digesters in the US [18]. One

of the most prevalent large-scale applications of anaerobic

digestion in the US is in landfills where anaerobic conditions

dominate the operational timeline [21]. However, as of June

2012, 594 of more than 1700 US landfills utilized biogas for en-

ergy production while the remainder flared biogas without

recovering biogas energy [22]. The active landfill projects lead to

generation of over 1800 MW of equivalent energy [15].

Compared to several other countries around the world,

anaerobic digestion has been relatively under-utilized in the

US for a variety of economic and technical reasons. These

include traditionally low energy and/or fuel prices, lack of

governmental incentives for implementing new anaerobic

power plants, the need for abundant and suitable land for site

development facilities and disposal of residuals, the need to

provide high quality reliable heat for the process to achieve

acceptable or commercially viable conversion efficiencies,

and the reputation of anaerobic processes as odor-generating

and difficult to operate [23e26]. The purpose of this paper is to

show through amodel system that when part of an integrated

system, anaerobic digestion can be a powerful resource for

wastemanagement and energy extraction. Specifically, in this

paper, a thermodynamic energy balance for amodel system is

presented demonstrating that anaerobic processing of waste

for harvesting both methane and ammonia as multiple fuel

sources in contrast to methane alone can provide an addi-

tional avenue to a net increase in extracted usable energy

from waste processing.

1.2. Inorganic nitrogen mitigation and removal

The environmental advantages of in-vessel anaerobic di-

gesters include stabilization of biochemical oxygen demand,

generation of biogas, production of digestate as a soil

amendment, and reduction of the environmental footprint

associated with land-filling [15,27,28]. However, several envi-

ronmental concerns as discussed below dictate post-

treatment steps needed for the digestate produced during

anaerobic digestion of organic feedstocks [29,30]. Of particular

concern is ammonia which is toxic to aquatic organisms,

causes eutrophication, and exerts oxygen demand in surface

waters [31]. Further, processes to remove ammoniaenitrogen

from aqueous effluent can require energy-intensive treatment

with large reaction vessels and long holding times [32e34].

Theammonia that accumulates inanaerobicdigesters exists

in two forms, ammonium ion ðNHþ4 Þ and free ammonia (NH3),

and is in equilibrium in aqueous systems (Equation (1)) [17].

NHþ44NH3 þHþ (1)

Total ammonia nitrogen (TAN) is the sum of NHþ4 and NH3

expressed as total N on a mass basis. The ratio of NH3-N to

NHþ4 �N in an aqueous system is governed by pH and tem-

perature (Equation (2)) [17].

½NH3 �N� ¼ ½TAN��1þ ½Hþ�

Ka

� (2)

where NH3-N is the free ammonia nitrogen concentration and

Ka is the temperature dependent dissociation coefficient for

Equation (2). TAN accumulates in digesters when proteins,

urea, nucleic acids, and other nitrogen-containing com-

pounds degrade, and its concentration must be controlled by

removal or by altering feedstock carbon to nitrogen (C:N) ra-

tios to prevent inhibition of the microbial process by higher

concentrations of free ammonia [35,36].

Anaerobic digestate is frequently used as a soil amend-

ment. However, application of anaerobic digestate to land as a

soil amendment must be carefully managed to avoid release

of excess nitrogen to surface waters, infiltration to ground

water, and the atmosphere. Particularly affected by these

problems are swine, poultry, and dairy operations, where land

application of digestate is an important disposal route [37e43].

Ma et al. (2005), for example, estimated that for Tompkins

County, NY, USAwith a total dairy herd of 9500 approximately

20,000 acres of suitable land would be needed to house di-

gesters and solids/liquids handling systems and to provide a

land sink for the resulting digestate [24]. With increasing

population pressures throughout the world, demand for such

large land resources can pose a significant problem for post-

processing of N or ammonia-rich waste feedstock. In addi-

tion, large domestic WWTP digesters located in metropolitan

environments are often at considerable distances from suit-

able land (>10 km). Consequently, these WWTPs must treat

ammonia onsite via nitrification/denitrification or haul

nitrogen-rich digestate to distant land sinks for disposal [24]

causing challenges for energy efficient waste management.

Anaerobic digester supernatant currently recycled to the

influent of some WWTPs may account for as much as 30% of

the incoming nitrogen loading to the facility [44] and also

constitutes a substantial regulatory concern and energy sink

[45]. Ammonia contained in leachate is also an important

factor controlling the long-term monitoring and post-closure

concerns of MSW landfills [46].

Conventional biological nutrient removal that combines

nitrification and denitrification requires long solids retention

b i om a s s a n d b i o e n e r g y 5 6 ( 2 0 1 3 ) 4 9 3e5 0 5 495

times (on the order of several hours to days) and energy-

intensive aeration to accommodate nitrifying bacteria

[32e34]. Further, denitrification mediated by heterotrophic

bacteria may divert carbonaceous substrates from methane

generation in digesters or require external electron donor

addition in wastewater applications [47]. Combined processes

of partial nitrification of ammonium to nitrite followed by

denitrification of nitrite (e.g., Canon/Sharon Anammox pro-

cesses [47e49]) have been developed to reduce energy and

oxygen demands and thus eliminate the need for external

electron donor addition [34]. However, these processes do not

eliminate the treatment energy demands completely nor do

they allow for the extraction of ammonia as a potential fuel.

Ammonia is a valuable industrial and agricultural chemical

used to produce fertilizer, solvents, cleaning agents, and re-

frigerants [50]. Furthermore, ammonia has been proposed as a

potential feedstock for hydrogen [51] due to the high density of

hydrogen per unitmass or volume of ammonia, and it can also

be directly harvested for energy conversion in a direct-

ammonia fuel cell [52e55]. Chemical routes for synthesis of

ammonia tend to be energy and material intensive [56] and

bio-ammonia as a sustainable fuel source is therefore

receiving increased interest for a variety of applications [57].

As discussed above, ammonia has a high-density of hydrogen

per unit volume on a weight basis of source material

(w0.18 g H2/g NH3), and compares favorably to other materials

used for hydrogen storage [58]. However, high temperature

(w800e900 �C) is required for thermal reforming to generate

hydrogen from ammonia i.e. energy input is needed to harvest

hydrogen as a fuel.

This paper evaluates whether ammonia liberated biologi-

cally (bio-ammonia) during anaerobic digestion could be har-

vested via stripping [59,60] and utilized as a source of hydrogen

as part of a coupled Anaerobic Digester-Bioammonia to

Hydrogen (ADBH) system (Fig. 1), and be a beneficial operational

approach in addition to harvesting methane from biogas.

Fig. 1 e Model anaerobic digester used for developing the

theoretical model for analysis. This system is referred to as

the anaerobic digester for bioammonia to hydrogen

(ADBH). The schematic shows flows of different streams

with details on each stream tabulated in Table 1. The

dotted line around the system represents the control

surface for thermodynamic analyses.

Ammonia-stripping has been used for treating animal waste

slurries [61e63], landfill leachate [64] and fertilizer plant wastes

[65]; however, it has not been extensively studied as a means

of ammonia removal from digester effluents [66]. Stripping

has been shown to reduce ammonia in effluents to less than

10 mg NH3-N/L [67]. Recovered ammonia gas could therefore

become the reforming fuel for catalytic reforming as shown in

Equation (3).

2NH3(g) / N2(g) þ 3H2(g) (3)

Thus, for an anaerobic digester producing biogas contain-

ing methane and discharging digestate rich in TAN, the in-

clusion of an ammonia recovery and reforming system to

generate hydrogen could allow additional biofuel or provide

an alternate route to harvesting an important industrial

chemical in itself. Consequently, the specific purpose of this

paper is to develop a conceptual model of an ADBH system

generating usable energy by harvesting multiple fuel source

streams in biogas and validate this concept model through a

thermodynamic energy balance based on the first law for

feedstocks of varying C:N ratios. Therefore, this paper (1) es-

tablishes a theoretical design scheme for an integrated system

to carry out anaerobic digestion and ammonia recovery to

demonstrate a quantifiable increase in overall energy gener-

ation from waste, (2) characterizes the energy demands and

energy production by focusing on two fuel sources in the

forms of methane and hydrogen, (3) considers different cases

of energy recovery within the integrated system to improve

the overall operation efficiency of the system from a net en-

ergy output perspective, and (4) identifies areas for further

scientific and engineering research needed to produce a net-

positive energy ADBH system. As discussed above, the en-

ergy balance estimates theoretical energy inputs and outputs

based on the first law of thermodynamics analyses and does

not account for process entropy changes.

2. Theoretical model framework

2.1. Model ADBH system description

The model ADBH system shown in Fig. 1 includes an anaer-

obic digester that stabilizes waste, produces biogas containing

methane and carbon dioxide as the main constituents, and

discharges digestate containing TAN. Further, the system

utilizes a solideliquid separator to concentrate the solid con-

tent in the digestate and produces liquid leachate containing

TAN. In addition, two pH shift reactors were included, to first

increase the pH of the leachate for converting TAN to NH3,

then later to neutralize the pH for recycle back to the digester.

After the leachate pH has been increased, gaseous ammonia is

recovered in a stripper. Finally, a combustion-based heat

source uses a fraction of the methane generated by the

digester as an energy source for ammonia reforming to pro-

duce hydrogen as an additional fuel. Descriptions of the con-

ceptual ADBH system flows indicated by arrows in Fig. 1 are

summarized in Table 1.

Stream 1, the solid organic waste flow, and Stream 2, the

influent additional liquid flow are the input streams to the

Table 1 e Theoretical anaerobic digester-bioammonia tohydrogen (ADBH) system flow descriptions andconstraints.

Number Type Phase Components

1 Influent Solid Organic waste feedstock

2 Influent Liquid

<1% TS

Aqueous stream

(TAN ¼ 200 mg/L)

3 Internal Slurry

10% TS

Digestate

4 Effluent Semi-solid

99% TS

Digestate solids with

CaCO3 precipitate

5 Internal Liquid

<1% TS

Digestate leachate

ðTAN ¼ NHþ4ðaqÞ þNH3ðaqÞÞ

6 Internal Liquid

<1% TS

Digestate leachate

(TAN ¼ NH3 (aq))

7 Recycle Liquid Digestate leachate with

N removed

8 Internal Liquid

<1% TS

Digestate leachate with

N removed (alkaline)

9 Recycle Liquid Digestate leachate with

N removed

10 Internal Gas NH3 (g)

11 Internal Gas Digester biogas (CH4, CO2)

12 Recycle Liquid

<1% TS

Digestate leachate with N

removed (neutral)

13 Internal Gas Concentrated biogas (CH4, CO2)

14 Effluent Liquid

<1% TS

Digestate leachate with N

removed (neutral)

15 Internal Gas Concentrated biogas (CH4, CO2)

16 Effluent Gas Effluent concentrated biogas

(CH4, CO2)

17 Influent Gas Air

18 Effluent Gas H2/N2

19 Effluent Gas Exhaust (CO2, H2O)

Table 2 e Variable input values for analysis of C:N ratioimpact on system energy accumulation.

Variable description Input value

C:N ratio Variable from 3

to 136 (g C/g N)

Dry-solids mass flow 1000 kg TS/h

Fraction of feedstock degraded 0.80

Moisture content (wt.% water) 90.0%

Aqueous TAN loading (in Stream 2) 200 mg NHþ4 �N=L

Percent recycle of Stream 8 65.0%

Ambient-digester temperature difference 40.0 �CInternal efficiency 0.35


anaerobic digester and control all individual system compo-

nents and reactor sizes, aswell as the internalmass flows, and

effluent mass flows. Stream 2 is controlled by the specified

operating moisture content in the digester, and the internal

liquid recycle flow (Stream 12). The inclusion of Stream 2 also

allows for treatment of aqueous flows containing TAN. The

volumetric or mass flow and composition of the influent

streamswas systematically varied to allow different scenarios

to be evaluated, as discussed later. In the following sections,

each system component is quantitatively described with a

discussion of net incoming or outgoing energy flows.

2.2. Anaerobic digester component

The anaerobic digester component consists of the anaerobic

reactor; incoming organic solid waste as Stream 1, a liquid

stream for achieving the desired operating moisture content

as Stream 2, liquid recycle Stream 12 formaximumusable fuel

recovery, outgoing biogas Stream 11, and the outgoing diges-

tate slurry from the anaerobic digester to the solideliquid

separator as Stream 3. In addition, heat and mechanical and/

or electrical power inputs to maintain digester temperature

and facilitate mixing were also included.

Influent Stream 1 for the digester is characterized by the

organic waste input quantified as a dry solids feedstock mass

flow rate, and is assigned both a specific carbon to nitrogen

(C:N) ratio (g C/g N) and a degradable organic fraction assumed

to be 80%, a commonly used fraction [17]. Additionally, the

reactor moisture content, the [TAN] concentration of the

additional liquid stream, and the internal leachate recycle

flow following ammonia stripping (Stream 12) were also

specified (Table 2).

Amodel feedstockmolecularmake-up based on a survey of

existing literature was designated to facilitate computation of

the methane, carbon dioxide, and TAN produced during

anaerobic digestion. The TAN of the digestate at pH 7 would

consist of 3.8% NH3 and 96.2% NHþ4 at an operational tem-

perature of 55 �C (Equation (2)). The digester is assumed to

operate under thermophilic conditions (55 �C) rather than

mesophilic conditions (37 �C) because a separate analysis

found that even minimal increases in methane generation

under thermophilic conditions were sufficient to offset any

additional heating demands, and could potentially net more

total fuel [67a]. Although no such methane production en-

hancements were assumed for analyses presented here,

maintaining conservative estimates of our system’s perfor-

mance, additional heating requirements were assessed for

operation at 55 �C. An additional advantage of thermophilic

operation over mesophilic operation is that the liberation of

ammonia from biomass during hydrolysis may proceed to a

greater extent and themicrobial communitymay have greater

tolerance to ammonia [68]. Two separate molecular feedstock

formulas were considered. First, the conversion of the

degradable fraction of the organic matter to methane and

carbon dioxide was via Equation (4) [19].

C6H10O4 þ 1.5H2O / 3.25CH4 þ 2.75CO2 (4)

Organic compounds containing nitrogen were modeled as

C2nH5nO2nNn to allow themolar amounts of nitrogen to vary in

response to the C:N ratio of the feedstock. A simple protei-

nogenic amino acid (glycine) was used as the empirical

structural basis for nitrogen containing species as it had the

greatest C:N ratio, and based on reported stoichiometry it was

taken that conversion of the organic nitrogen fraction to

methane proceeded via Equation (5) [69].

C2nH5nO2nNnþ0:5nH2OþHþ/0:75nCH4þ1:25nCO2þnNHþ4 (5)

The overall stoichiometry of the methanogenic breakdown of

the feedstock was therefore the sum of Equations (4) and (5) to

yield Equation (6).


Cð6þ2nÞHð10þ5nÞOð4þ2nÞNn þ�1:5þ 0:5n

�H2Oþ nHþ/

�3:25þ 0:75n

�� CH4 þ

�2:75þ 1:25n

�CO2 þ nNHþ

4

(6)

where n, the stoichiometric amount of nitrogen contained in

the feedstock formula (mol N/mol feedstock), was computed

from the C:N ratio of the influent feedstock by Equation (7).

n ¼ 6$ðAWCarbonÞðC : NÞratio$

�AWNitrogen

�� 2$ðAWCarbonÞ(7)

From Equations (6) and (7) it can be seen that there are 6moles

C per mol of the organic portion of the feedstock and 2 addi-

tional moles of C per mol of feedstock for each mole N added,

AWCarbon is the atomic weight of C (12 g C/mol C), AWNitrogen is

the atomic weight of nitrogen (14 g N/mol N) and (C:N)ratio is

the carbon to nitrogen ratio (g C/g N) contained in the desig-

nated feedstock.

To fully account for carbon and nitrogen flows in the

model, microbial growth was also included [70] in a manner

similar to other waste treatment mass balance models [71].

The net fraction of the feedstock (C(6þ2n)H(10þ5n)O(4þ2n)Nn)

electron equivalents incorporated into new microbial

biomass, fs, was calculated according to Equation (8) [17,70,72].

fs ¼ f 0s $

�1þ �

1� fd�$b$T

1þ b$T

�(8)

where f 0s is the theoretical fraction of feedstock electrons used

for synthesis and was taken to be 0.05 [72]; fd is the fraction of

the microbial biomass that is biodegradable and was taken as

0.80; b, the decay rate was 0.03 d�1 [17]; and T, the retention

time was 20 days. The net fraction of electrons used for en-

ergy, fe, was fe¼ 1� fs [17,70,72]. Incorporating fs and fe and cell

synthesis modified the overall digester reaction stoichiometry

(Equation (9)) [17] to

CaHbOcNd þ�2aþ d� c� 0:45efs � 0:25efe

�H2O/

�0:125efe

�CH4

þ �a� c� 0:2efs � 0:125efe

�CO2 þ

�0:05efs

�C5H7O2N

þ �d� 0:05efs

�NHþ

4 þ �d� 0:05efs

�HCO�

3

(9)

where a (6þ 2n), b (10þ 5n), c (4þ 2n) and d (n) correspond to the

feedstockmaterial (C(6þ2n)H(10þ5n)O(4þ2n)Nn), e¼ 4aþ b� 2c� 3d,

and themicrobial biomass was assumed to have themolecular

formula C5H7O2N, which controlled the amount of N incorpo-

rated during synthesis of new cells [17]. The TAN concentration

in the digestate under differing C:N ratios was equal to the

NHþ4 �N component of Equation (9). Further, the methane

generated from each test condition was determined from

Equation (9).

Calculations for external heating required to raise the

temperature of the influent material to 55 �C and also to

maintain the digester at 55 �C are described in Section S1 of

the Supplementary Materials. The power required for me-

chanically mixing the digester, which was assumed to be a

continuously stirred tank reactor (CSTR), is described in Sec-

tion S2 of the Supplementary Materials. The solidseliquid

separation unit, handling digestate flow from the digester,

was assumed to be a gravity separation unit operating at 95%

efficiency [73]. The unit size was a function of the influent

digestate total solids content (%TS), the volumetric flow rate,

the average assumed particle size (0.5 mm) [73], and the

specified solids removal efficiency (95%) [73]. Flow through the

unit was assumed to be driven by the hydrostatic pressure

drop between the digester and the first pH-shift reactor.

2.3. Ammonia gas recovery component

The ammonia gas recovery system consisted of two pH-shift

reactors, an aqueous ammonia stripper, and Streams 5e14.

The NHþ4 in the digestate leachate (Stream 5) was converted to

NH3 in the first pH shift reactor where calcium hydroxide

(lime) was added to increase the pH from 7 to 11 (Equations

(10) and (11)).

CaðOHÞ2/Ca2þðaqÞ þ 2OH�

ðaqÞ�DH ¼ �16:7 kJ=mol

�(10)

NHþ4ðaqÞ þOH�

ðaqÞ/NH3ðaqÞ þH2O�DH ¼ �366:6 kJ=mol

�(11)

Heat generated from reactions described by Equations (10) and

(11) is sufficient to increase the temperature of the aqueous

stream entering the stripper to the solution boiling point from

the digester discharge temperature of 55 �C. NH3(aq) contained

in Stream 6 is then stripped to the gas phase in the ammonia

stripper using steam, produced from Stream 7, as the strip-

ping gas. The alkaline, TAN reduced digestate leachate

(Stream 8) flows to the second pH shift reactor and the pH is

neutralized using the carbon dioxide in the biogas, Stream 11

(Equation (12)).

Ca2þ þ 2OH� þCO2/CaCO3 þH2O�DH¼�97:10 kJ=mol

�(12)

This neutralization process removes carbon dioxide from the

biogas stream and converts it to a dissolved form in the

aqueous phase, subsequently producing a concentrated

biogas stream with higher methane content than coming

directly from the digester (Stream 13). A finite energy input is

required to operate the ammonia stripper and to homogenize

the reactants in the pH-shift reactors; however, excess pro-

cess heat could be recovered from the reactions described by

Equations (10) and (12). This is included in themodel based on

the enthalpies of each reaction as discussed below, which

assume that these reactions go forward to completion. The

reactions are assumed to reach completion based on available

data (the concentrations of CaCO3 at pH 11 andOH� at pH 7 are

each less than 10�7 M respectively), and this assumption is

provided to validate the reaction enthalpy values utilized.

However, the model accounts for reaction equilibriums to

rectify stoichiometry and quantify mass balances. Thus, the

calculated heat values correspond to the predicted equilib-

rium of each reaction for which kinetic and equilibrium data

has been previously reported.

Elston and Karmarkar (2003) considered aqueous ammonia

stripping technologies for selective catalytic reduction appli-

cations (SCR) [67]. Comparing air and steam as stripping

media they identified a linear relationship for power re-

quirements for aqueous ammonia stripping (19 wt.% NH3(aq))

for a range of aqueousmass flows [67]. Based on their findings,

the power requirements for stripping ammonia from the

digestate leachate using either air or steam were estimated to

be 0.265 kJ/kg-aqueous mass treated (kJ/kg-aq.) or 1.574 kJ/kg-

aq., respectively [67].


Heat produced during each of the pH shifts was determined

using the standard enthalpy of formation (DHf) for the reactants

and products [74,75]. Heat from these exothermic reactionswas

assumed to be captured and re-used with an internal heat

transfer efficiency of 35% [76]. In this work, the energy input for

operating the mechanical components required optimizing

this internal heat utilization, for example, through the use of

additional pumping to and from a heat exchange unit, was

not estimated. The sensitivity of this latter assumption for

influencing the overall energy balance was assessed as part of

the energy integration cases (see Section 2.7).

Mixing of pH-shift reactors was assumed to occur via ve-

locity gradient mixing with retention times of 30 s [77] and the

power requirementwas determined as described in Section S3

of the Supplementary Materials.

2.4. Ammonia reforming component

The ammonia reforming component consisted of two sepa-

rate processes. The first was a combustion-based heat source

to heat a catalyst within a reformer where the incoming NH3(g)

(Stream 10) is catalytically pyrolyzed to yield N2 and H2

(Stream 18). A fraction of the concentrated methane biogas

(Stream 13) is diverted from the splitter to the combustor

(Stream 15), combined with air (Stream 17), and burned, pro-

ducing combustion exhaust (Stream 19). The ammonia

reforming reaction shown in Equation (13) is endothermic,

requiring heat to proceed.

2NH3ðgÞ/N2ðgÞ þ 3H2ðgÞ�DH ¼ 46 kJ=mol

�(13)

The reaction at 850 �C is approximately 90% efficient [78], and

the air stream was assumed to be at 15 �C, thus requiring

heating of 13.4 kJ/mol-NH3 converted for Equation (13). The

amount of biogas diverted to the combustor is dependent on

themethane content of the biogas, the energy demand for the

system, NH3(g) flow rate (Stream 10), and the heating value of

methane, assumed to be 50 MJ/kg (lower heating value of

methane for combustion) [79].

For all reactions listed, a single-step reaction mechanism

without reaction intermediates or kinetic considerations is

assumed. While the analytical model presented in this paper

could be optimized further, it nevertheless provides several

valuable insights into developing anaerobic digesters as a

potential waste processing system to generate multiple fuel

streams.

2.5. Stream transfer pressure-drop and pumpingrequirements

Power demands resulting from movement of internal gas and

aqueous flows were estimated using mass and energy bal-

ances for individual streams as described in Section S4 of the

Supplementary Materials. The associated power as a function

of the required pressure drop and flow for each internal

stream was assessed based on specific process configurations

as case-studies. Values for a specific process configuration or

stream were input variables in the model allowing different

configurations and operating conditions to be tested analyti-

cally. Although not a fluid stream, the energy of transporting

the feedstock (Stream 1) into the digester was calculated by

determining the work required to raise its mass to the height

of the digester assuming the energy utilization was 35% effi-

cient. The dimensions of the model system components were

determined as described in Section S4 of the Supplementary

Materials.

2.6. Mass and energy balance model

Material and energy flows for the conceptual system were a

function of the input flows (Table 2), which were systemati-

cally varied to estimate a system-wide energy balance for a

range of input values as discussed later. Further, mass flows of

C and N were tracked through the system, and C and N as a

percentage of the total C and N entering the system were

computed for each component flow. The modeled outputs

were a function of the system inputs (e.g. input flow rate, size,

geometry, retention times, etc.), which were held constant

while the feedstock C:N ratio was varied (Table 2). Results of

the energy balance were expressed either as a function of the

C:N ratio or as a function of the corresponding TAN as

described both in the text and accompanying figures. To

further evaluate the system, the methane plus hydrogen from

ammonia reforming was classified as ‘net usable energy’ as

these two materials would act as fuel sources. The energy

consumed by the ADBH system was normalized to the

methane production potential for the feedstock at each C:N

ratio as determined from Equation (9). It should be noted that

while the harvesting of bioammonia presents other avenues

for useful resources either as direct fuel or an important in-

dustrial chemical, those benefits are not accounted for in the

‘net usable energy’ estimates discussed in this work. The

normalization to the methane production potential was cho-

sen since methane is the most commonly and widely har-

vested fuel source from anaerobic biogas and provides a ready

reference for comparison. Therefore, the equivalent usable

energy was calculated as the heat generated during the

complete burn of the total fuel (methane and hydrogen) dur-

ing complete, single-step oxidation.

2.7. Model application to different case scenarios

Three test operating cases were examined for comparison

(Table S1). For all cases, identical inputs (Table 2) were used,

but the modeled outputs were calculated based on different

assumptions as discussed next. For Case 1, the heat generated

from chemical reactions and the hydrostatic pressure avail-

able from each reactor was assumed to be recaptured at 35%

efficiency to offset heating and pumping requirements else-

where in the process. For Case 2, the same assumptions were

made as for Case 1 with an additional assumption that the

system utilized the N2/H2 gas stream (Stream 18) as the

stripping gas, reducing the stripping power requirements

from 1.574 to 0.265 kJ/kg-aq. In Case 3, no internal energy re-

covery/integration was included.

The thermodynamic energy balance model was then used

to identify the effect of the feedstock C:N ratio on the overall

energy balance for Cases 1, 2, and 3 (Table S1) under steady-

state operation. A first law thermodynamic analysis was

conducted to estimate the theoretical maxima for the energy


output of the envisioned ADBH system to identify conditions

where net energy output exceeds energy input and potentially

presents a viable operating case. Other aspects of process

viability, such as economics and feasibility of the design to

account for irreversible (entropy) losses were not included in

the present analysis, but the currentmodel can be extended to

enable further work to include a second law analysis and an

economic analysis to demonstrate the viability of developing

such systems for fuel generation from anaerobic waste

processing.

Fig. 3 e ADBH mass percent flux of carbon with feedstock

C:N ratio of 3.0. Thicker arrows represent streams with

greater mass flows while the non-bold text displays the

breakdown percentages of the bold category directly above

it. Numbers in parentheses correspond to stream numbers

presented in Fig. 1.

3. Results and discussion

The energy balance of the ADBH system was determined for

feedstock C:N ratios (g C/g N) between 3.0 and 136, presenting

the lower and upper bounds for reasonable waste feedstock as

a starting point. The relationship between the feedstock C:N

ratio and corresponding digester TAN concentration is pre-

sented in Fig. 2. Higher C:N ratioswere not considered because

above 136, the system nitrogen, even combined with the an

aqueous nitrogen loading of 100 mg TAN/L, was too low to

support microbial cell growth as predicted by Reaction (5).

Similarly C:N ratios below 3 caused digester TAN to rise above

10,000 mg/L, which would severely limit the viability of

anaerobic microorganisms [35,36]. Next a discussion of the

three specific test cases is presented.

3.1. Mass balances on C and N

The effect of system energy integration as tested by comparing

Cases 1, 2, and 3 (see discussion below) did not change overall

system mass fluxes of C and N. Thus, Figs. 3 and 4 show the

mass fluxes for C and N, respectively, for a C:N ratio of 3 as a

representative example. The greatest mass percent C in the

effluent from the ADBH system was associated with carbonate

(32.4%) and methane (31.1%). The remainder was associated

with digestate (residual feedstock and biomass, 22.3%) and

Fig. 2 e Comparison of resulting digester TAN

concentration and feedstock carbon to nitrogen (C:N) ratio.

The limits of the C:N ratios were chosen on ability to

support microbial growth and viability.

carbon dioxide gas (14.2%). Note that the calcium carbonate

precipitate generated in the pH shift reactors accounted for

9.9% of the carbon flow. The greatest mass percent N in the

effluent was associated with nitrogen gas (61.1%), and the

remainder was associated with digestate (residual feedstock

and biomass, 38.9%).

3.2. Case 1 assessment (internal heat and pressuredifferentials recovered)

3.2.1. Fuel, heat, and energy analysisAs the C:N ratio decreases from 136 to 3 (i.e., the organic ni-

trogen content increases with fixed carbon content), the

operating TAN concentration in the digester increases from

Fig. 4 e ADBH mass percent flux of nitrogen with

feedstock C:N ratio of 3.0. Thicker arrows correspond to

streams with greater mass flows. Numbers in parentheses

correspond to stream numbers presented in Fig. 1.

Fig. 5 e Net system energy generation attributable to

methane and hydrogen respectively as a function of

steady-state digester TAN concentration (TAN is a function

of feedstock C:N ratio, as shown in Fig. 2).


20 mg/L to 10,000 mg/L. The model predicts that as the oper-

ating TAN concentration in the digester increases, more

hydrogen is recovered from the ammonia reformer and more

heat is recovered from the pH-shift reactors. Biofuel energy

fromhydrogen increases from 0 to 2.5MJ-biofuel/kg-feedstock

and heat output from the pH shift reactors increases from 0.64

to 1.02 MJ-heat/kg-feedstock.

The heat energy required to operate the digester (to in-

crease inlet feed flow temperature from 15 �C and to maintain

the digester at 55 �C) was not offset by the heat generated in

the pH-shift reactors for TAN concentrations below 7500mg/L

(i.e., C:N ratios > 4.3). Thus, the system requires external heat

for operation up to a TAN concentration of 7500 mg/L. This

external heat input decreases as the operating TAN concen-

tration in the digester increases, and the need for external

heating becomesminimal at TAN concentrations of 7500mg/L

because the heat produced in the pH-shift reactors is a func-

tion of the TAN concentration.

The mass flow rate of ammonia (kg/h) recovered is pro-

portional to TAN concentration increases and the power input

required to strip the dissolved ammonia from the aqueous

phase. However, since the stripping power required is a

function of the total aqueous mass flow (10,047 kg/h), the

marginal increase in the ammonia to be recovered has a

minimal impact (<1%) on the stripping power demand.

The model shows that the methane diverted to facilitate

heating and reforming in the combustor also increases as the

amount of ammonia gas treated increases. In return for the

potentially usable methane energy that is diverted to the

combustor to drive the ammonia reformer, an increasing

amount of hydrogen energy is recovered. The equivalent en-

ergy from hydrogen increases as the operating TAN concen-

tration in the digester increases between 20 and 10,000 mg/L

(between C:N ratios of 136 and 3.0). As a result, the net system

bio-fuel energy generation (i.e., the sum of the equivalent

energy associated with the biogas comprising primarily

methane and hydrogen through reforming of ammonia) de-

creases as a function of the operating TAN concentration

(Fig. 5). The relative contribution of methane and hydrogen to

the net energy shows that methane is decreasing and

hydrogen is increasing as the C:N ratio decreases (Fig. 5). Note

that the sum of equivalent methane and hydrogen energy at

10,000 mg/L TAN (C:N ratio 136:1) is less than the equivalent

methane energy at 20 mg/L TAN (C:N ratio 3:1) (Fig. 5). Thus

overall, as the C:N ratio decreases, net usable energy pro-

duction decreases. This result is a product of the reaction

stoichiometry shown in Reaction (5), where it is observed that

the total amount of methane that can be produced from a

fixed mass of organic substrate decreases as the C:N ratio

decreases. This occurs because as the C:N ratio decreases,

more of the model substrate described in Equation (6) is

incorporated. As a result of the reaction stoichiometry

(Equation (9)), the moles of C available for producing methane

decrease. Therefore, it would appear that the ADBH system is

not a viable technology for harvesting fuel streams other than

methane, which in itself would be a major finding. However,

as shown next this is not the complete picture.

To further evaluate the ADBH system, the energy gener-

ated and consumed by the process was normalized to the

methane production alone from the feedstock at each C:N

ratio (Fig. 6). The normalization permits a direct comparison

in the net usable system energy in contrast to the energy that

can be potentially generated by extracting and consuming

methane only as a fuel. For higher C:N ratios, the energy

produced approaches the total methane potential of the sub-

strate (100% in Fig. 6b). This is based on the stoichiometry

from Equation (9) and the defined fraction of the degradable

portion defined in Table 2.

As the C:N ratio decreases (TAN concentration in the

digester increases), and ammonia stripping with hydrogen

recovery is utilized, some methane is diverted to the

combustor to power the ammonia reforming process. Thus,

themethane recovered for external fuel use decreases relative

to the total available methane. However, the normalized

result shows that the reduction in methane via diversion to

heat the combustor is compensated by the hydrogen energy

produced (Fig. 6b). Results above 100% indicate that the total

recoverable energy from the ADBH process recovering both

hydrogen and methane exceeds the potential recoverable

energy from methane alone.

The thermodynamic analysis shows that the recovered

equivalent hydrogen energy is not always sufficient to offset

the system operation energy requirements. The amount of

energy consumed by the system to generate (via the digester),

recover, and reform ammonia is greater than the equivalent

hydrogen energy generated for TAN concentrations below

approximately 8500 mg/L (C:N < 3.7). However, the amount of

energy consumed by the system to simply recover and reform

ammonia, excluding energy needed to operate the digester, is

greater than the equivalent hydrogen energy generated for

TAN concentrations below approximately 1000 mg/L

(C:N < 31). Thus, operation of the ADBH system may not be

Fig. 6 e a) Potential methane production rate as a function

of feedstock stoichiometry controlled by feedstock C:N

ratio; b) Energy from methane and hydrogen as a

percentage of the potential methane production versus

operating digester TAN concentration (TAN is a function of

feedstock C:N ratio).


favorable for feedstock C:N ratios greater than 31 (corresponds

to digester TAN less than 1000mg/L). In general, operating the

ADBH system at TAN concentrations above 1000 mg/L (C:N

less than 31) improves the overall energy available by asmuch

as 1.8% at 1000 mg/L and 26.3% at 10,000 mg/L compared to

anaerobic digestion with methane recovery alone.

3.2.2. Influent feedstock effects on digester TAN and biogasqualityThe effect of the influent feedstock C:N ratio on digester TAN

and effluent biogas quality was analyzed by assuming and

holding constant one set of specific operating conditions

(Table 2). Although the operating TAN concentration in the

digester is used as a reference for the energy outputs in this

analysis, it is the C:N ratio of the dry solids flow that controls

TAN since other inputs are held constant. The designated

waste stream inputs and operating parameters were based on

typical design values [73] that contain a solid fraction of 10% or

less of the total incoming volume. Hence the specific heating

properties for the waste stream were approximated to be that

of liquid water in agreement with previously established

methods [73]. The operating TAN concentration was used as a

reference because high TAN concentrations (above

5000e6000 mg/L) have been reported to decrease methano-

genesis as a result of toxicity and inhibition of the microbial

community [80]. A C:N ratio between 6.8 and 5.6 would

correspond to digester TAN concentration greater than

5000e6000 mg/L based on the input parameters to the model

used in this paper. One caveat should be noted with respect to

the energy balance reported here. The presentmodel does not

explicitly evaluate process trade-offs to account for the rela-

tionship between the feedstock C:N ratio and the operating

TAN concentration in the digester which are a function of the

solids flow rate, recycle flow rate, set moisture content in the

digester, retention time, and digester aspect ratio. While such

a detailed system optimization will be no doubt useful it is

beyond the scope of the current effort. Specifically, for an in-

dividual C:N ratio (even for low ratios, C:N ¼ 1e3), the design

characteristics of the system could be adjusted to maintain a

viable operating TAN concentration (and corresponding free

ammonia concentration, the form of TAN thought to impose

toxicity [81,82]) in the digester such that TAN levels would not

become inhibitory. It is expected that this work will spur

further debate in the scientific community to design and build

ADBH systems accounting for complete system operation and

optimization.

The C:N ratio of the feedstock also affects the system

biogas composition by controlling the amount of carbon di-

oxide removed from Stream 11 in the second pH-shift reactor.

As the C:N ratio increases over the considered range, the

molar percent of carbon dioxide removed in the second pH

shift reactor rose from 22.0% to 48.2%. The biogas composition

in Stream 11 thus varied from 54.3% to 62.8% methane as the

C:N ratio was decreased. For feedstock C:N ratios of 136 and 3,

the methane content leaving the second pH-shift (Stream 13)

was thus predicted to be 64.3% and 80.7% methane, respec-

tively. This result presents another tradeoff to be considered.

If purified methane were the desired downstream product,

decreasing the C:N ratio of the influent solids material stream

can be seen as a means of improving downstream biogas

methane enrichment. Such an approach would impact the

energy balance for a system seeking to produce a nearly pure

methane stream with no carbon dioxide.

3.2.3. Power and pressure drop requirementsRequired pressure drops to power fluid flow, compensating for

frictional energy losses along the length of the pipe, were

included in the model (see Section S4 and Table S1 of the

Supplementary Materials) based on well-established pipe

flow models [83]. The magnitude of the frictional losses and

corresponding power required to maintain the calculated

pressure drop was almost exclusively a function of the spec-

ified influent feedstock mass flow rate, which was held con-

stant throughout the analysis. Thus, it was not unexpected

that the total power requirements for pumping (sum of the

hydrostatic and frictional power requirements), mixing, and

Fig. 7 e Normalized total equivalent energy generated by

ADBH system as a function of the operating TAN


ammonia stripping increased by less than 2% (from 86.4 to

87.5 kW) as the C:N ratio decreased from 136 to 3. A compar-

ison of the relative power demands for pumping, mixing, and

ammonia stripping showed that the mixing in the digester

and pH-shift reactors required the most power (Table 3). The

sum of the power to sustain mixing in the digester and pH-

shift reactors, maintain the pressure drop and facilitate

ammonia stripping from the aqueous stream was two orders

ofmagnitude smaller than the equivalent energy generated as

fuel (methane rich biogas and ammonia reformed to

hydrogen) and heat. Subtracting the total power required for

these demands from the net energy equivalent generated

consequently had a relatively minor impact on the ADBH

system energy generation potential.

The overall process fuel and heat energy generation as a

function of the influent solids C:N ratio is shown in Fig. 7. The

normalized output indicates that above an operating digester

TAN concentration of approximately 2000 mg/L, the total

biofuel energy (CH4 and H2) output is greater than could be

expected to be produced from anaerobic digestion (CH4) alone

providing theoretical evidence for an increase in net usable

energy for implementing an integrated ADBH system.

concentration in the digester (TAN is a function of

feedstock C:N ratio). The solid line represents the expected

externally usable energy produced by anaerobic digestion

alone.

3.3. Effects of system energy integration: comparison ofcases 1, 2, and 3

The effects of various energy integration scenarios (Table S3)

were assessed by comparing the fraction of the energy

generated diverted to power the integrated process as a

function of the operating digester TAN concentration. The

percent of the energy required increased as the influent

feedstock C:N ratio decreased (corresponding TAN concen-

trations in the digester increased) for all cases because

pumping and mixing requirements downstream from the

digester were a function of the mass flow of ammonia. Addi-

tionally, the energy content of the feedstock decreased as the

C:N ratio decreased, allowing less methane to be generated

(Fig. 6a) from approximately 18 kmol/h for a C:N ratio of 136

decreasing to nearly 7 kmol/h for a C:N ratio of 3. The total

percentage of energy diverted remained under 35% for all

cases and feedstock C:N ratios. The ADBH system operated

under the Case 1 and Case 2 scenarios did not generate greater

amounts of methane or hydrogen than when operated under

the Case 3 scenario, but as expected, less biofuel (methane

Table 3 eADBH fractional power requirements: demandsfor pumping, mixing and ammonia stripping, as well asrelative requirements of mixing in the digester and pH-shift reactors.

Process Fractional power requirement

Mixing 83%

Digester 42%

pH-Shift 1 29%

pH-Shift 2 29%

Stripping 15%

Pumping 2%

and/or hydrogen) was diverted to provide energy needed by

the ADBH system.

Case 3 had the highest fraction (23e35%) of its biofuel being

diverted to power various systemdemands for all feedstock C:N

ratios because available energy integration was not considered

in this case, i.e., heat generated and available hydrostatic

pressures to facilitate stream flow were not utilized. With no

energy recovery, the system losses were 23 to 35% of the total

energy generated (fraction of methane and hydrogen).

Recovery of heat and available hydrostatic pressure were

considered for integration in Case 1 and Case 2. Case 2 differed

from Case 1 in that it also utilized a fraction of the available

hydrogen and nitrogen stream (Stream 18) as the stripping gas

as opposed to steam alone. For Case 1, liquid vaporization and

stripping required 1.574 kJ/kg-aq., whereas for Case 2, using

the available gas stream eliminated the need for vaporization

to produce steam, and ammonia stripping thus required only

0.265 kJ/kg-aq, improving the amount of energy recovered

nearly 7 times. When internally produced heat and pressure

differentials were recovered (Case 1), system losses were

reduced to between 8 and 17% in contrast to a high of 35% for

no energy recovery. The energy required for ammonia strip-

ping in Case 2 assumed that the hydrogen and nitrogen

stream (Stream 18) could strip ammonia from the liquid

stream as effectively as air. Using steam as the stripping gas

allows ammonia to be concentrated by allowing steam to be

selectively removed via condensation. It is likely that the

minimal (<1%) reduction in energy use observed by utilizing a

fraction of the nitrogen and hydrogen stream for stripping

would be offset by the need to re-concentrate the diluted

reforming gas stream.


3.4. Summary, implications and areas for furtherinvestigation

The ADBH system energy balance remained positive (gener-

ating net usable energy in contrast to energy consumed) while

digesting biomass, removing aqueous ammonia, and reform-

ing ammonia gas to produce hydrogen for all feedstock C:N

ratios considered, but the energy balance favored C:N ratios

lower than 31. The model did not account for microbial

sensitivity to aqueous TAN concentrations, entropic losses,

and, unless otherwise specified, assumed 1-step reactions and

Arrhenius kinetics. Feedstocks C:N ratios (below 4.0) might be

better managed by employing a separate hydrolysis-

fermentor reactor upstream from the digester. Ammonia

release during hydrolysis occurs more rapidly than other

anaerobic processes in the digester [68], and operating a

separate hydrolysis reactor could potentially reduce nitrogen

loading in the downstream digester. This could prevent

ammonia toxicity and improve digester stability andmethane

generation, while still allowing bioammonia to hydrogen

conversion.

As the model and discussion above illustrates, more

research needs to be conducted before full-scale ADBH sys-

tems could be pursued. The cost for constructing and oper-

ating ADBH systems versus the relative benefits of energy

generation and nitrogen removal need to be assessed and

evaluated in the context of alternative methods of disposing

of TAN. Since aqueous nitrogen species are typically a concern

for digester effluents and may require significant subsequent

treatment, the energy and economic savings of using ADBH

systems versus conventional digestion and downstream ni-

trogen mitigation processes can be important. Biological

mediation of aqueous nitrogen species are typically aerobic

processes that require organic substrate, and do not offer the

prospect of energy recovery. In fact, aeration typically is the

greatest energy sink at wastewater treatment plants and the

organic substrate required could be viewed as substrate that is

diverted from methane generating anaerobic processes.

Finally, though not considered explicitly in this work, har-

vesting of ammonia as a direct fuel source or chemical source

can also be pursued, providing additional avenues to manage

waste streams and eliminating the need to divert methane for

internal processes as in the present ADBH system.

4. Conclusions

The total equivalent energy generated by the ADBH system is

greater than the energy consumed to operate the system for

the conditions considered over the entire range of C:N ratios

tested in contrast to extracting methane only as an energy

source. When the ammonia recovery and combustor systems

are coupled to the digester system (becoming the single

ADBH system), the equivalent energy from hydrogen alone is

not sufficient to offset system energy requirements with

digester operation when the feedstock C:N ratio is above 3.7

(<8500 mg/L TAN). Operating the ADBH system increased the

total energy recovery potential compared to the methane

potential for feedstock C:N ratios less than 31. Without

integrating the digester process energy, the ADBH system

only requires 23e34% of the total energy generated (fraction

of methane and hydrogen), and process integration can

reduce the system loss to between 8 and 17%. For a C:N ratio

of 3, the lowest value evaluated, the integrated ADBH process

produced 25.5% more energy than anaerobic digestion

generating methane alone. Finally, it is expected that this

analysis and design scheme will spur debate within the sci-

entific community, and hopefully lead to more detailed

optimization and future construction of integrated ADBH

systems for harvesting usable fuel feedstocks from waste

streams.

Acknowledgments

This project was funded in part by the Rutgers University

Academic Excellence Fund. The authors thank the Depart-

ment of Environmental Sciences and the New Jersey Water

Resources Research Institute at Rutgers University for partial

support as well as Defense Advanced Research Projects

Agency (DARPA) and the Samsung Group for personnel sup-

port. The authors also acknowledge support from the

Department of Mechanical Engineering at The Ohio State

University for support during preparation of the manuscript.

Shaurya Prakash would also like to thank Prof. Mark A.

Shannon at the University of Illinois for several insightful

discussions.

Appendix A. Supplementary material

Supplementary material associated with this article can be

found, in the online version, at http://dx.doi.org/10.1016/j.

biombioe.2013.05.024.

r e f e r e n c e s

[1] Guest JS, Skerlos SJ, Barnard JL, Beck MB, Daigger GT,Hilger H, et al. A new planning and design paradigm toachieve sustainable resource recovery from wastewater.Environ Sci Technol 2009;43:6126e30.

[2] Grant SB, Saphores J-D, Feldman D, Hamilton A, Fletcher T,Cook P, et al. Taking the “waste” out of “wastewater” forhuman water security and ecosystem sustainability. Science2012;337:681e5.

[3] Augedlo-Vera C, Leduc W, Mels A, Rijnaarts H. Harvestingurban resources towards more resilient cities. ResourConserv Recycl 2012;64:3e12.

[4] Iranpour R, Stenstrom M, Tchobanoglous G, Miller D,Wright J, Vossoughi M. Environmental engineering: energyvalue of replacing waste disposal with resource recovery.Science 1999;285:706e11.

[5] Wang X, Junxing L, Ren N-Q, Yu H-Q, Lee D-J, Guo X.Assessment of multiple sustainability demands forwastewater treatment alternatives: a refined evaluationscheme and case study. Environ Sci Technol2012;46:5542e9.

[6] McCarty PL, Bae J, Kim J. Domestic wastewater treatment as anet energy producer e can this be achieved? Environ SciTechnol 2011;45:7100e6.


[7] Nallathambi-Gunaseelan V. Anaerobic digestion of biomassfor methane production: a review. Biomass Bioenergy1997;1:83e114.

[8] Schink B. Anaerobic digestion: concepts, limits andperspectives. Water Sci Technol 2002;45:1e8.

[9] Nasir IM, Ghazi TM, Omar R. Production of biogas fromorganic solid wastes through anaerobic digestion: a review.Appl Microbiol Biotechnol 2012;95:321e9.

[10] Weiland P. Biogas production: current state andperspectives. Appl Microbiol Biotechnol 2010;85:849e60.

[11] Demirel B, Scherer P, Yenigun O, Onay T. Production ofmethane and hydrogen from biomass through conventionaland high-rate anaerobic digestion processes. Crit RevEnviron Sci Technol 2010;40:116e46.

[12] De Baere L, Mattheeuws B. Anaerobic digestion of MSW inEurope. BioCycle 2010;51:24.

[13] Hawkes DL. Review of full-scale anaerobic digestion inChina. Agric Wastes 1986;18:197e205.

[14] Ho M. Biogas bonanza for third world development. London,UK: Institute of Science in Society; 2005.

[15] Kellaher M. Anaerobic digestion outlook for MSW streams.Biocycle Energy 2007;48:51e5.

[16] Ortenblad H. Development of anaerobic digestion plants andprogress in selected countries. Herning, Denmark:EnergiGruppen Jylland; 2002.

[17] Rittmann BE, McCarty PL. Anaerobic treatment bymethanogenesis. Environmental biotechnology: principlesand applications. Boston, MA: McGraw Hill; 2001. p. 570e629.

[18] AgSTAR U, editor. Anaerobic digesters continue growth inU.S. livestock market. Washington D.C.: USEPA; 2011.

[19] Themelis NJ, Ulloa PA. Methane generation in landfills.Renew Energy 2007;32:1243e57.

[20] USEPA. Municipal wastewater treatment facilities. In:Combined heat and power partnership. Washington D.C.:USEPA; 2010.

[21] Reinhart DR, Townsend TG. Landfill bioreactor design andoperation. Boca Raton, FL (USA): CRC Press LLC; 1998.

[22] USEPA. Energy projects and candidate landfills. Landfillmethane outreach program. Washington DC: USEPA; 2012.

[23] Chen Y, Cheng JJ, Creamer KS. Inhibition of anaerobicdigestion processes: a review. Bioresour Technol2008;99:4044e64.

[24] Ma J, Scott NR, DeGloria SD, Lembo AJ. Siting analysis offarm-based centralized anaerobic digester systems fordistributed generation using GIS. Biomass Bioenergy2005;28:591e600.

[25] Matteson GC, Jenkins BM. Food and processing residues inCalifornia: resource assessment and potential for powergeneration. Bioresour Technol 2007;98:3098e105.

[26] Brown BB, Yiridoe EK, Gordon R. Impact of single versusmultiple policy options on the economic feasibility of biogasenergy production: swine and dairy operations in NovaScotia. Energy Policy 2007;35:4597e610.

[27] Chung SS, Poon CS. Evaluating waste managementalternatives by multi criteria approach. Resour ConservRecycl 1996;17:189e210.

[28] Murphy JD, Power N. A technical, economic, andenvironmental analysis of energy production fromnewspaper in Ireland. Waste Manag 2007;27:177e92.

[29] Van Horn HH, Hall MB. Agricultural and environmental issuesin the management of cattle manure. Agricultural uses of by-products and wastes. American Chemical Society; 1997.

[30] Chernicharo CA. Post-treatment options for the anaerobictreatment of domestic wastewater. Rev Environ SciBiotechnol 2006;5:73e92.

[31] Metcalf, Eddy. Constituents in wastewater. In:Tchobanoglous G, Burton F, Stensel HD, editors. Wastewaterengineering. Boston: McGraw Hill; 2003. p. 27e137.

[32] Metcalf, Eddy. Biological denitrification. In:Tchobanoglous G, Burton F, Stensel HD, editors. Wastewaterengineering: treatment and reuse. Boston: McGraw Hill; 2003.p. 619.

[33] Metcalf, Eddy. Aggregate organic constituents. In:Tchobanoglous G, Burton F, Stensel HD, editors. Wastewaterengineering: treatment and reuse. Boston: McGraw Hill; 2003.p. 87.

[34] Ahn YH. Sustainable nitrogen elimination biotechnologies: areview. Process Biochem 2006;41:1709e21.

[35] Chen Y, Cheng JJ, Creamer KS. Inhibition of anaerobicdigestion process: a review. Bioresour Technol2008;99:4044e64.

[36] Kayhanian M. Ammonia inhibition in high-solidsbiogasification: an overview of practical solutions. EnvironTechnol 1999;20:255e365.

[37] Clemens J, Trimborn M, Weiland P, Amon B. Mitigation ofgreenhouse gas emissions by anaerobic digestion of cattleslurry. Agric Ecosyst Environ 2006;112:171e7.

[38] Hansen KH, Angelidaki I, Ahring BK. Anaerobic digestion ofswine manure: inhibition by ammonia. Water Res1998;32:5e12.

[39] Freibauer A. Biogenic emissions of greenhouse gases fromEuropean agriculture. Eur J Agron 2003;19:135e60.

[40] Pechan Z, Knappova O, Petrovicova B, Adamec O. Anaerobicdigestion of poultry manure at high ammonium nitrogenconcentrations. Biol Wastes 1987;20:117e31.

[41] Rumburg B, Mount G. Atmospheric flux of ammonia fromsprinkler application of dairy waste. Atmos Environ2006;40:7246e58.

[42] Vanotti MB, Szogi AA, Hunt PG, Millner PD, Humenik FJ.Development of environmentally superior treatment systemto replace anaerobic swine lagoons in the USA. BioresourTechnol 2007;98:3184e94.

[43] Whitehead TR, Cotta MA. Isolation and identification ofhyper-ammonia producing bacteria from swine manurestorage pits. Curr Microbiol 2004;48:20e6.

[44] Bowden G. Fennell DE, editor. Nitrogen reflux at wastewatertreatment facilities. New York, NY: AECOM; 2008.

[45] Gander M, Jefferson B, Judd S. Aerobic MBRs for domesticwastewater treatment: a review with cost considerations.Sep Purif Technol 2000;18:119e30.

[46] Berge ND, Reinhart DR. The fate of nitrogen in bioreactorlandfills. Crit Rev Environ Sci Technol 2005;35:365e99.

[47] Jetten M, Schmid M, Schmidt I, Wubben MJGK. Improvednitrogen removal by application of new nitrogen-cyclebacteria. Rev Environ Sci Biotechnol 2002;1:51e63.

[48] Strous M, Fuerst J, Kramer E, Logemann S, Muyzer G, Pas-Schoonen KVD, et al. Missing lithotroph identified as newplanctomycete. Nature 1999;400:446e9.

[49] Hellinga C, Schellen AAJC, Mulder JW, Loosdrecht MV.Sharon process: an innovative method for nitrogen removalfrom ammonium-rich wastewater. Water Sci Technol1998;37:135e42.

[50] Ammonia. In: Bohnet M, Bellussi G, Bus J, Cornils B, Drauz K,Greim H, et al., editors. Ullmann’s encyclopedia of industrialchemistry. John Wiley & Sons, Inc.; 2010.

[51] Dunn-Rankin D, Leal EM, Walther DC. Personal powersystems. Prog Energy Combust Sci 2005;31:422e65.

[52] Maffei N, Pelletier L, Charland JP, McFarlan A. An intermediatetemperature direct ammonia fuel cell using a protonconducting electrolyte. J Power Sources 2005;140:264e7.

[53] Ma Q, Peng R, Lin Y, Gao J, Meng G. A high-performanceammonia-fueled solid oxide fuel cell. J Power Sources2006;161:95e8.

[54] Fournier GGM, Cumming IW, Hellgardt K. High performancedirect ammonia solid oxide fuel cell. J Power Sources2006;162:198e206.


[55] Meng G, Ma G, Ma Q, Peng R, Liu X. Ceramic membrane fuelcells based on solid proton electrolytes. Solid State Ionics2007;178:697e703.

[56] USDOE. Energy and environmental profile of the U.S.chemical industry. Agricultural chemicals: fertilisers.Washington DC: US Department of Energy; 2000 [chapter 5].

[57] Zamfirescu C, Dincer I. Using ammonia as a sustainable fuel.J Power Sources 2008;185:459e65.

[58] Klerke A, Christensen CH, Norskov JK, Vegge T. Ammonia forhydrogen storage: challenges and opportunities. J MaterChem 2008;18:2285e392.

[59] Air stripping for ammonia removal. In: WEF, editor. Design ofmunicipal wastewater treatment plants. Alexandria, VA:WEF; 1998.

[60] Sedlak R. Air stripping of ammonia. In: Sedlak R, editor.Phosphorus and nitrogen removal from municipalwastewater: principles and practice. Boca Raton, FL: CRCPress LLC; 1991.

[61] Gangagni Rao A, Sasi Kanth Reddy T, Surya Prakash S,Vanajakshi J, Joseph J, Jetty A, et al. Biomethanation ofpoultry litter leachate in UASB reactor coupled withammonia stripper for enhancement of overall performance.Bioresour Technol 2008;99:8679e84.

[62] Liao PH, Chen A, Lo KV. Removal of nitrogen from swinemanure wastewaters by ammonia stripping. BioresourTechnol 1995;54:17e20.

[63] Bonmatı A, Flotats X. Air stripping of ammonia frompig slurry:characterisation and feasibility as a pre- or post-treatment tomesophilic anaerobic digestion. Waste Manag 2003;23:261e72.

[64] Cheung KC, Chu LM, Wong MH. Ammonia stripping as apretreatment for landfill leachate. Water Air Soil Pollut1997;94:209e20.

[65] Minocha VK, Prabhakar AVS. Ammonia removal andrecovery from urea fertilizer plant waste. Environ TechnolLett 1988;9:655e64.

[66] Lei X, Sugiura N, Feng C, Maekawa T. Pretreatment ofanaerobic digestion effluent with ammonia stripping andbiogas purification. J Hazard Mater 2007;145:391e7.

[67] Elston JT, Karmarkar D. Aqueous ammonia strippingtechnology for SCR applications. Houston, TX: Electric Power;2003;[67a]Babson DM. Enhancing energy recovery from biomasswaste streams e from mega-landfills and biorefineries tomicrobial communities. Ph.D. dissertation, RutgersUniversity; 2010.

[68] Gallert C, Bauer S, Winter J. Effect of ammonia on theanaerobic degradation of protein by mesophilic andthermophilic biowaste population. Appl Microbiol Biotechnol1998;50:495e501.

[69] Ramsay IR, Pullammanappallil PC. Protein degradationduring anaerobic wastewater treatment: derivation ofstoichiometry. Biodegradation 2001;12:247e57.

[70] Rittmann BE, McCarty PL. Stoichiometry and bacterialenergetics. Environmental biotechnology: principles andapplications. Boston, MA: McGraw Hill; 2001. p. 126e61.

[71] Ekama GA. Using bioprocess stoichiometry to build a plant-wide mass balance based steady-state WWTP model. WaterRes 2009;43:2101e20.

[72] Rittmann BE, McCarty PL. Microbial kinetics. Environmentalbiotechnology: principles and applications. Boston, MA:McGraw Hill; 2001. p. 165e98.

[73] Metcalf, Eddy. Treatment, reuse, and disposal of solids andbiosolids (anaerobic digestion). In: Tchobanoglous G,Burton F, Stensel HD, editors. Wastewater engineering.Boston: McGraw Hill; 2003. p. 1505e46.

[74] Tester JW, Modell M. Standard enthalpy and Gibbs freeenergy of formation. In: Amundson NR, editor.Thermodynamics and its applications. 3rd ed. Upper SaddleRiver, NJ: Prentice Hall PTR; 1996. p. 291.

[75] Tester JW, Modell M. Estimating physical properties: groupcontribution methods for estimating pure componentproperties. In: Amundson NR, editor. Thermodynamics andits applications. 3rd ed. Upper Saddle River, NJ: Prentice HallPTR; 1996. p. 555e83.

[76] Neveu P, Castaing J. Solidegas chemical heat pumps: field ofapplication and performance of the internal heat of reactionrecovery process. Heat Recovery Syst CHP 1993;13:233e51.

[77] Metcalf, Eddy. Physical unit operations. In: Tchobanoglous G,Burton F, Stensel HD, editors. Wastewater engineering. 4thed. Boston: McGraw Hill; 2003.

[78] Ganley JC, Seebauer EG, Masel RI. Porous anodic aluminamicroreactors for production of hydrogen from ammonia.AIChE J 2004;50:829e34.

[79] NIST. NIST standard reference database number 69. NIST;2008. 2008 ed.

[80] Bhattacharya SK, Parkin GF. The effect of ammonia onmethane fermentation processes. Water Pollut Control Fed1989;61:55e9.

[81] Sprott GD, Patel GB. Ammonia toxicity in pure culturesof methanogenic bacteria. Syst Appl Microbiol1986;7:358e63.

[82] Gallert C, Winter J. Mesophilic and thermophilic anaerobicdigestion of source-separated organic waste: effect ofammonia on glucose degradation and methane production.Appl Microbiol Biotechnol 1997;48:405e10.

[83] Wilkes JO. Mass, energy, and momentum balances. In:Anzalone J, editor. Fluid mechanics for chemical engineers.Upper Saddle River, NJ: Prentice-Hall, Inc.; 1999.

anaerobic digestion for methane generation and ammonia … · anaerobic digestion for methane...

Documents