wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 4 7 5 8e4 7 6 8

Avai lab le at www.sc iencedi rect .com

journa l homepage : www.e lsev ier . com/ loca te /wat res

Net energy production associated with pathogen inactivationduring mesophilic and thermophilic anaerobic digestion ofsewage sludge

Christopher Ziemba, Jordan Peccia*

Department of Chemical and Environmental Engineering, Yale University, New Haven, CT 06520, USA

a r t i c l e i n f o

Article history:

Received 14 March 2011

Accepted 15 June 2011

Available online 24 June 2011

Keywords:

Biosolids

Pathogens

Energy

Biogas

Methane

Reactivation

* Corresponding author. Tel.: þ962 203 432 4E-mail address: jordan.peccia@yale.edu (

0043-1354/$ e see front matter ª 2011 Publidoi:10.1016/j.watres.2011.06.014

a b s t r a c t

The potential for anaerobic digester energy production must be balanced with the

sustainability of reusing the resultant biosolids for land application. Mesophilic, thermo-

philic, temperature-phased, and high temperature (60 or 70 �C) batch pre-treatment

digester configurations have been systematically evaluated for net energy production

and pathogen inactivation potential. Energy input requirements and net energy production

were modeled for each digester scheme. First-order inactivation rate coefficients for

Escherichia coli, Enterococcus faecalis and bacteriophage MS-2 were measured at each digester

temperature and full-scale pathogen inactivation performance was estimated for each

indicator organism and each digester configuration.

Inactivation rates were found to increase dramatically at temperatures above 55 �C.

Modeling full-scale performance using retention times based on U.S. EPA time and

temperature constraints predicts a 1e2 log inactivation in mesophilic treatment, and a 2e5

log inactivation in 50e55 �C thermophilic and temperature-phased treatments. Incorpo-

rating a 60 or 70 �C batch pre-treatment phase resulted in dramatically higher potency,

achieving MS-2 inactivation of 14 and 16 logs respectively, and complete inactivation (over

100 log reduction) of E. coli and E. faecalis. For temperatures less than 70 �C, viability staining

of thermally-treated E. coli showed significantly reduced inactivation relative to standard

culture enumeration. Due to shorter residence times in thermophilic reactors, the net

energy production for all digesters was similar (less than 20% difference) with the 60 or

70 �C batch treatment configurations producing the most net energy and the mesophilic

treatment producing the least. Incorporating a 60 or 70 �C pre-treatment phase can

dramatically increase pathogen inactivation performance without decreasing net energy

capture from anaerobic digestion. Energy consumption is not a significant barrier against

improving the pathogen quality of biosolids.

ª 2011 Published by Elsevier Ltd.

1. Introduction centralized sewage collection systems, activated sludge-based

More than 7 million dry tons of sewage sludge are produced

annually in the U.S. (Beecher et al., 2007). This figure is

expected to increase as more communities move to

385; fax: þ962 203 432 438J. Peccia).shed by Elsevier Ltd.

nutrient removal processes become more prevalent, pop-

ulations served by sewers grow, and anaerobic digestion is

developed as a renewable energy source. The U.S. EPA

encourages the treatment and beneficial reuse of stabilized

7.

wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 4 7 5 8e4 7 6 8 4759

sewage sludge (biosolids) as a solution to the high costs and

environmental impacts of incineration and land filling.

Approximately 55% of treated sewage sludge in the U.S. is

reused, primarily as a soil conditioning product and fertilizer

(Beecher et al., 2007).

While the agricultural benefits of land applying biosolids

are well documented (Evanylo et al., 2008; Khaleel et al., 1981),

there is widespread concern over pathogen exposure to resi-

dents in communities that surround land applications sites

(Lewis and Gattie, 2002; NRC, 2002). Pathogen inactivation

during anaerobic digestion is an integral component for

ensuring the safety and sustainability of biosolids reuse. With

limited information from which to estimate risk, the U.S.

EPA’s part 503 biosolids regulations were based on using best

available treatment technology to address pathogen concerns

(USEPA, 1999). These regulations established class A and class

B treatment standards and prescribed stabilization methods

necessary to meet each designation. Class A biosolids must

undergo an EPA-approved treatment process yielding fecal

coliform concentrations less than 1000 colony forming units

(CFU) per dry gram, or less than 3 most probable number

(MPN) Salmonella sp. in 4 dry grams. The resulting class A

biosolids can be sold and utilized without restriction. Class B

pathogen reduction goals are less stringent, requiring fecal

coliform concentrations less than 2 � 106 CFU or MPN per dry

gram. Class B biosolids still contain human pathogens

following treatment, therefore site restrictions and reductions

in vector attraction are also required to reduce pathogen

exposure to the public. The most common treatment tech-

nology for meeting class B requirements is anaerobic diges-

tion operating in the mesophilic range of 35e40 �C. Achievingclass A standards through digestion usually involves

increasing digester temperature to between 50 and 55 �C for

thermophilic anaerobic digestion (TAD). Temperature-phased

anaerobic digestion (TPAD) is a hybrid process which

commonly features a shorter TAD phase, for hydrolysis,

pathogen inactivation, and sometimes acetogenesis, followed

by a mesophilic anaerobic digestion (MAD) phase.

These current anaerobic digestion configurations have

known limitations for inactivating pathogens. MAD inactiva-

tion efficiency is limited and only results in one or two log

removal of fecal indicators in full-scale operation over a 15e40

day residence time (Gantzer et al., 2001; Guzman et al., 2007;

Pedersen, 1981). Although log fecal coliform reduction is

significantly greater during traditional thermophilic

processes, (50 or 55 �C) and potentially on the order of four

logs, increasing evidence suggests that TAD may not be reli-

able at permanently inactivating some bacterial pathogen

indicators. TAD processes may induce a viable but non-

culturable (VBNC) condition, from which some bacteria may

later recover. Such reactivation behavior has been demon-

strated during high-speed centrifugal dewatering of thermo-

philically digested biosolids in both Escherichia coli and fecal

enterococci (Higgins et al., 2007; Qi et al., 2007; Sahlstrom

et al., 2004; Viau and Peccia, 2009a). Reactivation has not

been reported in pasteurized biosolids. Mechanistic evidence

of ribosomal unfolding at different temperatures (Lee and

Kaletunc, 2002; Mackey et al., 1991) suggest that fecal coli-

form bacteria can be fully inactivated at temperatures at and

above approximately 60 �C.

Redesigning digesters to employ an effective pathogen

inactivation mechanism that does not significantly increase

energy consumption is essential to ensuring the safe and

economical reuse of biosolids and the development of biogas

as a sustainable alternative fuel. We hypothesize that incor-

porating a 60 �C or higher initial phase into an anaerobic

digestion process will significantly improve pathogen inacti-

vation performance over mesophilic or thermophilic treat-

ment, and that the energy efficiencies produced by a shorter

residence time and more effective hydrolysis will allow for

this greater inactivation without impacting net energy

production value. This hypothesis has been evaluated by

estimating pathogen indicator inactivation rate coefficients in

sludge as a function of temperature, identifying common

(MAD, TAD, TPAD) and alternative digestion schemes (60 and

70 �C pre-treatment), and modeling the pathogen reduction

and net energy production of each scheme. Rather than

a strict focus on meeting current regulations, this research

seeks to understand the energy costs associated with

decreasing the pathogen load in land applied biosolids.

2. Materials and methods

2.1. Selection of indicator organisms

Escherichia coli was chosen as a test organism because it is

a member of the fecal coliform class, upon which the U.S. EPA

503 regulations for biosolids pathogen quality are predomi-

nantly based. Enterococcus faecalis is a representative member

of the fecal enterococci group, which are gram-positive and

have been shown to be more resistant to temperature inacti-

vation than fecal coliforms (Viau and Peccia, 2009a). MS-2 is

a commonly studied male-specific (Fþ) coliphage. A compel-

ling case has been presented in the literature that male-

specific coliphages have value as an indicator of fecal

contamination and pathogenic virus inactivation due to their

similarity to human enteric viruses in terms of structure and

persistence through treatment (Funderburg and Sorber, 1985;

Havelaar et al., 1993; Nappier et al., 2006).

2.2. Batch temperature inactivation experimentalprocedure

Batch testing was conducted in 125 ml crimp-top serum

bottles (Wheaton, Millville, NJ, USA) containing 74.5 ml of

mesophilically digested sewage sludge, adjusted to 6% solids

in phosphate buffered saline solution (PBS, 0.14MNaCl, 0.01M

phosphate, and 0.003 M KCl, American Bioanalytical, Natick,

MA). The digested sludge was obtained locally from a munic-

ipal wastewater treatment plant that utilized activated sludge

treatment, MAD stabilization and belt filter press dewatering.

Sludge characteristics are typical and presented in

Supplementary Data Table S-1. Sludges were prepared by

autoclaving for 30 min, homogenizing in a blender for 20 min

and adjusting pH to 7.5. Bottles were capped, purged with

nitrogen and acclimated to experimental temperature in

either an incubator (T ¼ 37 �C) or a water bath (T ¼ 50, 55, 60 or

70 �C). Each reactor bottle was anaerobically spiked with

0.5 ml of one of three pathogen indicator organisms, E. coli, E.

wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 4 7 5 8e4 7 6 84760

faecalis or MS-2 bacteriophage. Well mixed conditions were

established initially by vigorous shaking and maintained by

orbital shaking at 40 RPM throughout the experiments. Reac-

tors were sampled periodically by 18-gauge needle-tipped

syringe, and cultures immediately enumerated.

Due to rapid inactivation rates and slow characteristic

mixing times for the 6% solids solution, inactivation experi-

ments for E. coli and E. faecalis at 60 and 70 �C were conducted

in a series of 1 ml syringes. Sludge media was processed as in

the serum bottle reactors described above, with the exception

that the inoculumwas added and the bottles well mixedwhile

at room temperature. The inoculated sludge media was then

distributed into 1 ml syringes, capped, and submerged in

a water bath at 60 or 70 �C. After waiting 60 s for the 60 �Creactors and 80 s for the 70 �C reactors to reach temperature

(heating times were independently verified), syringes were

individually removed at 5e10 s intervals and bacteria were

enumerated. All inactivation experimental conditions were

tested in duplicate. Log-transformed data were pooled and

first-order inactivation rate constants and associated stan-

dard errors of fit were estimated by least squares regression.

To determine the potential for VBNC behavior, E. coli batch

experiments were recreated as above, at 50, 55, 60, and 70 �Cusing PBS in place of sludge. Each bottlewas inoculatedwith E.

coli and incubated for times corresponding to w4 logs of

culture-based inactivation. Bottleswere cooled in a 25 �Cwater

bath and concentrations of viable cells were determined by

staining with 5-cyano-2,3-ditolyltetrazolium chloride (CTC,

SigmaeAldrich, St. Louis, MO). Viable staining procedure con-

sisted of incubating 100 ml of sample in the dark for 90 min at

37 �C with 1/100 strength TSB at a final concentration of 5 mM

CTC.Viable cell countswere standardized to total cell count, by

independently staining with SYTO-9 (Invitrogen, Carlsbad,

CA). Total cell countprocedure consistedof incubating100ml of

sample in the dark for 20min at room temperature with a final

SYTO-9 concentration of 5 mM. CTC and SYTO-9 stained cells

were each enumerated using standardmicroscopy procedures

and appropriate fluorescent filters (Hobbie et al., 1977). Each

temperature was tested in triplicate at a residence time cor-

responding to a 4 log loss of culturability. Each residence time

was estimated and confirmed by plate count, employing the

methods described below.

2.2.1. Inoculum preparationOvernight cultures of E. coli (ATCC#15597, American Type

Culture Collection, Manassas, VA) were grown at 37 �C in 5 ml

of tryptic soy broth (TSB, Difco Inc., Detroit, MI)). A 50 ml

aliquot of this overnight culture was reinoculated in 50 ml of

TSB and incubated at 37 �C underwellmixed conditions for 5 h

to reachmid-log phase. The bacteria werewashed three times

by repeated centrifugation at 5000 g for 10 min and resus-

pension in PBS, and finally suspended in 5 ml of PBS. Bacterial

concentration was determined by direct fluorescence

microscopy using 40, 6-diamidino-2-phenylindole (DAPI)

staining (SigmaeAldrich, St. Louis, MO) and standard counting

procedures (Hobbie et al., 1977). Preparation of E. faecalis

(ATCC#19433) was identical to that of E. coli, except brain heart

infusion broth (Himedia Inc., Mumbai, India) was used.

Themale-specific bacteriophageMS-2 (ATCC15597-B1)was

prepared using a modified double agar layer (DAL) method

(USEPA, 2001). In a hot water bath at 48 �C, 1 ml of MS-2 stock

and 1ml of log phaseE. coli (ATCC#15597)were added to 4ml of

molten 0.7% tryptic soy agar (TSA, Difco Inc., Detroit, MI)) in

a 30ml test tube. The test tubewas immediately removed from

thewater bath, gently rolled and poured onto a 10 cmdiameter

TSA plate. The DAL plates were allowed to cool to room

temperature, inverted and incubated at 37 �C for 24 h. To elute

phages, 2 ml of PBS was gently added to each plate and again

incubated at 37 �C for 1 h. The PBSwashwas then collected and

centrifuged at 5000 g for 10 min. The supernatant was passed

through a 0.45 mm syringe filter to remove bacteria (Whatman

Inc. Florham Park, NJ, USA), and MS-2 phages were concen-

trated using a centrifugal membrane (Millipore Inc. Billerica,

MA, USA). Final MS-2 concentration was determined by

repeating the DAL method on dilutions from the new stock

solution and counting the number of plaques in the E. coli lawn

for dilutions yielding between 30 and 300 plaques. Inoculum

concentrations were on the order of 109 CFU/ml for bacteria

and 1012 PFU/ml (plaque forming units) for MS-2.

2.2.2. Enumeration by culturingReactor samples were serially diluted in PBS to achieve

between 30 and 300 colony or plaque forming units per culture

plate. E. coli was plated on mFC agar (Difco Inc., Detroit, MI)

and incubated at 44.5 �C for 24 h (APHA et al. 2005). E. faecalis

was plated on mEI agar (Difco Inc., Detroit, MI) and incubated

at 41.5 �C for 48 h (USEPA, 2002). Enumeration of MS-2 was

conducted by the same DAL method described above, with

serial dilutions of 0.45 mm filtered reactor samples added to

the E. coli and molten TSA in place of the MS-2 Stock. All

plating was performed in duplicate.

2.3. Digester configurations

Table 1 presents seven digestion configurations selected to

represent commonly used mesophilic and thermophilic

digester temperatures, plus 60 �C and 70 �C proposed

pretreatment schemes. The selected residence times are

based on U.S. EPA part 503 class A and class B pathogen

regulations. Class B MAD treatment at 37 �C is modeled at the

U.S. EPA minimum of 15 days residence time (USEPA, 1994).

For digesters operating at or above 50 �C, with solids content

less than 7%, the EPA-mandated minimum residence times to

achieve class A standards are governed by a time and

temperature relationship (USEPA, 1994). Application of this

relationship yields a minimum residence time of 5 days in

a digester at 50 �C. TAD residence times are extended to 15

days to allow more complete solids conversion and to better

reflect the longer residence times often employed in practice

(Viau and Peccia, 2009b). The residence times in our 50 and

55 �C TPAD configurations are based on the 5 day minimum

requirement at 50 �C, followed by the standard 15 days in

MAD. At 60 and 70 �C, the EPA mandated minimum residence

times are approximately 5 and 0.5 h respectively for pathogen

inactivation. Table 1 presents a more conservative and inter-

nationally established residence time of 1 h for 70 �Cpasteurization, and the specified 5 h at 60 �C. Both 60 and 70 �Ctreatmentsmust be paired with amesophilic phase to achieve

acceptable volatile solids (VS) conversion and gas production.

The second phase is modeled as a 15 day MAD reactor.

Table 1 e Time and temperatures of the seven anaerobic digester configurations evaluated. MAD and TAD configurationsare single stage processes. (cstr) denotes completely mixed stirred tank reactor configuration.

Configuration Phase 1 Phase 2

Temperature (�C) Residence Time (days) Temperature (�C) Residence Time (days)

MAD 37 15 (cstr)

TAD 50 50 15 (cstr)

TAD 55 55 15 (cstr)

TPAD 50 50 5 (cstr) 37 15 (cstr)

TPAD 55 55 5 (cstr) 37 15 (cstr)

60 batch þ MAD 60 0.208 (5 h batch) 37 15 (cstr)

70 batch þ MAD 70 0.042 (1 h batch) 37 15 (cstr)

wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 4 7 5 8e4 7 6 8 4761

Each digestion scheme has been configured to incorporate

heat exchangers to capture waste heat and reduce total

energy costs, a common trend in modern digester design

(Greer, 2007; Zupancic and Ros, 2003). The efficiency (the

difference between the final and initial temperatures in the

hot stream divided by the difference in initial temperatures of

the cold and hot streams) of counter-flow heat exchangers are

typically 50% (Kepp et al., 2000). For single-phase MAD and

TAD configurations listed in Table 1, heat from the digester

effluent sludge is transferred to the digester influent sludge by

a counter-flowheat exchanger. In TPAD, 60 �C batch, and 70 �Cbatch configurations, incoming sludge receives heat from two

sets of counter-flow heat exchangers. Here, influent sludge is

first preheated by captured heat from the effluent of the

second (MAD) phase and once again heated by the waste heat

from sludge leaving the first (thermophilic CSTR, 60 or 70 �Cbatch) phase.

2.4. Energy balances

The amount of energy produced by an anaerobic digester per

metric ton of wet sludge, is defined as the per ton energy

content of the biogas producedminus the per ton energy input

demands to operate the digester. This input demand includes

energy required to heat sludge to digester operating temper-

ature and the energy required to compensate for heat losses

during operation. Secondary costs such as stirring and

secondary products such as biological heat generation during

digestion are relatively insignificant (Lubken et al., 2007) and

are omitted from the net energy calculation. All analysis

assumes a 6% solids content.

2.4.1. Heat-up energy demandThe amount of heat (kWh) required per wet metric ton

(1000 kg) to heat-up sludge is the difference between the initial

and desired temperatures multiplied by the specific heat

capacity of 6% solids sludge, 1.117 � 103 kWh/kg �C (Metcalf

and Eddy, 2003). The initial temperature of sludge flowing

into a digester is the temperature at the previous source plus

temperature gained through heat exchanger recovery.

Incoming sludge to each digester configuration is initially

assumed to be 15.6 �C (Metcalf and Eddy, 2003).

2.4.2. Heat loss from reactorsThe rate of heat loss from a reactor, ( _q reactor, W) is described in

equation (1) as the sum of heat loss rates through the floor,

walls and roof.

_q reacter ¼all surfaces

�Usurface Asurface ðTreacter � ToutÞ

�(1)

X

The heat loss through each surface is the product of the

overall heat transfer coefficient through the surface (Usurface,

W/m2/ �C), the surface area (Asurface, m2) and the difference in

inside (Treacter, �C) and outside (Tout, �C) surface temperatures.

For the purpose of defining surface areas, plausible reactor

geometries have been selected to accommodate a theoretical

flow of 1.2� 102metric tons (0.45million gallons) of wet sludge

per day (Metcalf and Eddy, 2003), which corresponds to

a wastewater flow of approximately 3 � 107 L (9 million

gallons) per day. Digesters are conventionally cylindrical in

shape, drawing to a point at the bottom. For 15 day residence

time reactors, dimensions are set at 18 m in diameter, 6 m

deep at the sides and 9 m deep in the middle. The dimensions

of 5 day residence time reactors preserve the same ratio

between the diameter, side depth andmid depth, scaled down

to 1/3 volume. Wall construction for 15 and 5 day residence

time reactors is set at 0.3 m thick concrete in contact with air,

resulting in an overall heat transfer coefficient of 4.9 W/m2 �C

(USEPA, 1979). The floor is also 0.3 m thick concrete, in contact

with dry earth, resulting in an overall heat transfer coefficient

of 0.34 W/m2 �C, (USEPA, 1979). Selecting a 0.225 m thick fixed

concrete cover achieves an overall heat transfer coefficient of

3.3 W/m2 �C (USEPA, 1979). Representative external tempera-

ture values are 11.5 �C for soil and 8.6 �C for air.

In batch reactor configurations at 60 or 70 �C, the sludge is

treated using three batch reactors. The dimensions are 2.9 m

diameter, 1.9 m side depth and 2.6 mmid depth for each 60 �Creactor and 1.5 m diameter, 1.6 m side depth and 2.1 m mid

depth for each 70 �C reactor. Both sets of batch reactors are

constructed using 10 mm thick steel walls, floor and roof (Le

et al., 2002), each in contact with air, resulting in an overall

heat transfer coefficient, Usurface, of 5.167 W/m2 ��C (USEPA,

1979). It is assumed that batch reactors operate constantly

and operate at capacity. We assume 15 min fill and empty

times for 1 h residence time reactors and 30min fill and empty

times for 5 h residence time reactors. Heat losses from batch

reactors during filling and emptying phases are conservatively

modeled to be identical to heat losses during full operation.

The overall amount of heat loss per wet ton of sludge for

a specific reactor configuration (qreacter, Wh/ton treated

sludge) is depicted in Eq. (2), and determined by multiplying

the heat loss rate ( _qreacter, W, Eq. (1)) by the residence time of

interest (qreactor) and then dividing by the reactor volume

(Vreactor, L) and the density of sludge (rsludge, kg/L). The density

of 6% sludge is 1.01 kg/L (Metcalf and Eddy, 2003).

wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 4 7 5 8e4 7 6 84762

qreacter ¼

( Pall surfaces

�Usurface Asurface ðTreactor � ToutÞ

�)qreactor

Vreactor rsludge

(2)

2.4.3. Biogas productionBiogas production was estimated using a generalized organic

waste fermentation equation that models digested sludge as

C10H19O3N and utilizes CO2 as the terminal electron acceptor,

Eq. (3) (Rittmann and McCarty, 2001).

C10H19O3Nþ �18� 22:5fs � 12:5fe

�H2O/

�6:25fe

�CH4

þ�9� 10fs � 6:25fe

�CO2 þ

�2:5fs

�C5H7O2Nþ �

1� 2:5fs�NHþ

4

þ�1þ 2:5fs

�HCO�

3 (3)

Thenet fractionof sludge organicmatter converted to cellmass

is represented as fs and the net fraction of this organic matter

that is utilized for cellular energy is represented by fe Values of

f s¼ 0.07 (and f e¼ 0.93) were estimated by previously published

relationships based on a mass balance for volatile solids in

a CSTR (Rittmann and McCarty, 2001) and using the initial

fraction converted into cells fos ¼ 0:11, the biodegradable frac-

tion of cellular biomass f d ¼ 0.8, an endogenous decay rate

b¼ 0.05 day�1, and the steady state CSTR residence timeӨ of 15

days.

A common VS conversion value of 56% was assumed for

mesophilic digestion at 15 days (Metcalf and Eddy, 2003).

Operating single phase digesters at thermophilic tempera-

tures or the presence of an initial thermophilic phase typically

improves hydrolysis and increases the bioavailability of

digestible material. Increases over MAD in VS destruction due

to either thermophilic digestion or the presence of a thermo-

philic phase have been reported to range from 7% to 11% (Ge

et al., 2010; Salsabil et al., 2010; Shimp et al., 2003). The TAD,

TPAD, and 60 and 70 �C pretreatment cases presented here are

modeled conservatively as achieving a 7% increase in VS

conversion (63% overall) over MAD treatment alone.

The raw energy value of methane (kWh) produced per ton

of digested sludge under each digester configuration can be

calculated by multiplying together the solids content (kg/kg),

the fraction VS of TS (kg/kg), the VS conversion efficiency of

the reactor (kg/kg), the number of moles of methane produced

per mole of converted sludge (mole/mole), the lower heating

value of methane (LHVCH4 ; kWh=mole), and divided by the

molecular weight of C10H19O3N (Msludge, g/mol), Eq. (4).

Incoming sewage sludge is modeled as 6% solids (Shimp et al.,

2003), with a 70% VS content (Metcalf and Eddy, 2003). The

lower heating value of methane is 0.223 kWh/mole (Perry and

Green, 1984).

Raw Energy ProducedMetric Ton Digested Sludge

¼

�% SolidsContent

��VSTS

��VS ConversionEfficiency %

��mole CH4

mole VS

�ðLHVCH4

Þ�Msludge

�� 106gmetric ton

� (4)

Finally, extracting energy value from biogas requires use-

specified purification of the gas. This model compares

energy values in terms of heat and therefore does not include

efficiencies associated with electricity generation, thus the

final value of the biogas includes only a 2% loss in total biogas

energy value for removal of water content by refrigeration

(Krich et al., 2005) and a 12% reduction from the raw energy

value due to combustion efficiency (Bekkering et al., 2010).

3. Results

3.1. Pathogen indicator inactivation kinetics determinedin anaerobic batch reactors

Inactivation rate coefficients for E. coli, E. faecalisandMS-2,were

measured in anaerobic batch reactors at 37, 50, 55, 60, and 70 �C(Fig. 1a,b,c). Inactivation profiles for each organism and

temperature are presented in Supplementary Data figures S-1,

S-2 and S-3. Increasing reactor temperature was found to

increase inactivation rate coefficients in each test organismand

the magnitude of this increase was much greater at tempera-

tures above 50 �C. Inactivation rate coefficients for E. coli and E.

faecalis in the 50e55 �C range are not statistically different

( p < 0.05) within each temperature, achieving 1.4 h�1 for E. coli

vs. 1.0h�1 forE. faecalisat50 �Cand6.8h�1 forE. colivs. 6.7h�1 for

E. faecalisat 55 �C.These rate coefficientswere 1e3 times greater

and statistically different ( p < 0.05) than that of MS-2 at 50 and

55 �C. Increasing temperature from55 to 60 �C yields a dramatic

increase in bacterial inactivation rate coefficient (significant at

p < 0.0001), and a smaller increase for MS-2. The inactivation

rate coefficient for E. coli is 94 times greater at 60 �C than at 55 �C(650 h�1) and 25 times greater in E. faecalis (177 h�1). The rate

coefficient for MS-2, in contrast, only doubled to 6 h�1. For

bacterial inactivation at 70 �C, it was determined that the 80 s

required to heat up the sludge to 70 �Cwas sufficiently lethal to

achieve complete inactivation at our limit of detection. There-

fore, no precise data are reported for E. coli or E. faecalis inacti-

vationat70 �C.Wecanconservativelyestimate inactivationrate

coefficients based on the log reduction from the seeded

concentration (typically 109 CFU/ml) to the method limits of

detection (2 � 102 CFU/ml) over the 80 s interval to be greater

than 106 h�1 for both bacteria. MS-2 inactivation at 70 �Cincreasedmore modestly to 36 h�1. At each digestion tempera-

ture MS-2 inactivation rate constants are significantly lower

( p< 0.05) than the correspondingvalues for E. coliand E. faecalis.

The inactivation kinetics obtained from batch reactor

testing have been inserted into first-order mass balance

models for CSTR or batch reactors to predict inactivation

performance for the seven residence time and reactor

temperature configurations listed in Table 1. MAD (37 �C)inactivation performance is limited to 1.1 log in E. coli, 1.6 log

in E. faecalis and 0.7 log in MS-2 (Fig. 2). Reductions in

Fig. 1 e First-order inactivation rate coefficients for E. coli,

E. faecalis and MS-2 as a function of temperature.

Experiments were carried out in batch reactors under

anaerobic conditions in autoclaved MAD biosolids at 6%

solids content and pH 7.5. Rate coefficients are

extrapolated by regression of the culturable concentration

measurements at each experimental temperature as

a function of time. Error bars represent standard error.

Insets in Figs. 1a, b and c display subsets of the full plot in

order to observe the impact of temperature on inactivation

rate coefficients below 60 or 70 �C.

Fig. 2 e Estimated digester performance for E. coli, E.

faecalis and MS-2 as log reductions in culturable

concentrations under various treatment schemes. Arrows

on E. coli and E. faecalis bars for 60 and 70 �C batch plus

MAD indicate dramatically higher log reduction values, in

excess of 100 log.

wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 4 7 5 8e4 7 6 8 4763

thermophilic single stage and phased digesters ranged from 2

to 4 logs in TAD at 50 or 55 �C and TPADat 50 �C configurations.

TPAD treatment at 55 �C displayed slightly greater inactiva-

tion with 4.0 log in E. coli, 4.5 log in E. faecalis and 3.2 log in MS-

2. Projected inactivation at 60 �C is significantly greater than in

MAD, TAD or TPAD configurations, resulting in inactivation

predictions of 13.7 log for MS-2 and greater than 100 log for E.

coli and E. faecalis. At 70 �C, E. coli and E. faecalis are again

projected to have greater than 100 log reduction and MS-2 is

predicted to be reduced by 16.4 log. Inactivation projections

for batch reactors do not include additional inactivation

occurring during filling and emptying phases.

Inactivation of E. coli at temperatures of 50, 55, 60 and 70 �Cwas also accessed by CTC viability staining and is presented in

Fig. 3. The initial viability in the E. coli inoculum was 68% of

total cells. After incubation at each temperature at times

standardized to a 4 log loss of culturability, losses in viability

were observed to be less than the losses in culturability in E.

coli. A reduction of 0.76 log viability was achieved at 50 �C, 1.31log at 55 �C and 2.14 log at 60 �C. At 70 �C the average loss of

viability was 2.93 log. The losses in viability at 50 and 55 �C are

significantly reduced from losses at 60 �C ( p < 0.001) and 70 �C( p < 0.01). The 70 �C loss of viability value of 2.93 log is not

significantly less ( p ¼ 0.063) than the 3.7 log loss exhibited in

autoclaved E. coli. The 70 �C and autoclaved E. coli reductions

are at the upper value of log reductions that can be observed

by microscopy.

3.2. Energy associated with inactivation

The total heating demand of each reactor has been divided by

the expected pathogen inactivation performance to show the

relative heating demands per log inactivation, Table 2. Total

heat demand is the sum of energy to heat the sludge up to the

digester temperature and the energy required to overcome

heat losses from the digester to maintain the desired

temperature. Due to the high inactivation rates at thermo-

philic temperatures, and the shorter residence times,

increasing the temperature decreases the amount of heating

energy required per log removal. Rankings of heat demand per

log removal are the following: MAD > TAD > TPAD>>60 and

70 �C pretreatment plus MAD.

Net energy output (kWhr/metric ton) is defined as the

amount of usable energy produced in biogas minus the total

heat demand. Each digestion configuration produces more

Temperature (°C)

50 55 60 70

Lo

g R

ed

uctio

n in

V

iab

le C

on

cen

tratio

n

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Fig. 3 e Log reduction of viable E. coli as determined by CTC

staining after batch temperature treatment at residence

times corresponding to w4 log loss of cultivability. Error

bars represent standard error of three independent

experiments.

wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 4 7 5 8e4 7 6 84764

energy than it consumes, and generally the amounts of energy

produced are similar across all configurations, with energy

produced in the most productive configuration only 19%

greater than the least (Fig. 4). Heating up sludge and heat

losses account for approximately 25% (range 19%e32%) of the

total energy produced in these digesters. The batch pretreat-

ment configurations produce themost net energywith both 60

and 70 �C pretreatment variations producing about 120 kWh

per wet metric ton of treated sludge.

4. Discussion

The results of this study reveal two important concepts for

operating anaerobic digesters to produce a pathogen-free,

sustainable biosolids product. First, batch temperature inac-

tivation studies revealed the dramatic increase in inactivation

potential of operating a digester at or above 60 �C. Second, thissignificantly greater pathogen reduction can be achieved

without a decrease in digester net energy production. The

Table 2 e Inactivation performance efficiency expressed as thesludge for a given reactor configuration divided by the log remconfiguration. Values for E. coli and E. faecalis batch plus MADestimates for inactivation rate coefficients and therefore prese

Configuration Heating Energy Demand(kWh/wet metric ton treated sludge)

MAD 27

TAD 50 42

TAD 55 47

TPAD 50 37

TPAD 55 39

60 batch þ MAD 31

70 batch þ MAD 29

information presented here demonstrates the feasibility and

the sustainability advantages of operating anaerobic digesters

at increased temperatures. This work is only an initial step.

Implementation at the full-scale must involve comprehensive

pilot testing and operational experience to ensure all chal-

lenges of digester performance (e.g. VS destruction, limited

foaming, inactivation of multiple types of human pathogens,

etc.) are met.

4.1. Inactivation kinetics

The major concern surrounding land applying class B

biosolids is exposure of workers and nearby residents to

infectious pathogens (NRC, 2002). Currently 39 of 50 U.S. states

have local or statewide restrictions on land application

(Beecher et al., 2007). Restrictions adopted typically include

the use of buffer zones between biosolids-applied land and

residential areas. The value of using set-back distances for

risk reduction however, is limited as aerosol transport studies

suggest that for most set-back distances (usually less than

50 m), there is less than 1 log reduction in infectious pathogen

exposure (Low et al., 2007). Multiple log reductions in path-

ogen content and potential risk, therefore, must be accom-

plished during sludge stabilization.

Temperatures above 55 �C have traditionally not been

considered in anaerobic digestion. Exceeding 55 �C in single

stage digesters may cause an imbalance between aceto-

genesis and methanogenesis and lead to instability or

digester failure. Incorporating 60 or 70 �C temperatures in the

first phase of multi-phase treatment, or as a batch-

pretreatment step, expands the range of operating tempera-

tures without negatively impacting process stability or solids

conversion. This study demonstrates a sharp increase in

bacterial first-order inactivation rate coefficients in pure

culture batch reactors at temperatures of 60 �C and above.

Such behavior is consistent with the onset of permanent

ribosome damage in E. coli occurring at temperatures in the

vicinity of 60 �C as determined by differential scanning

calorimetry (DSC) (Lee and Kaletunc, 2002; Mackey et al.,

1993). Unlike temperatures above 60 �C, DSC plots at 37, 50

and 55 �C do not reveal permanent conformational changes

in the structure of the cell, which would indicate effective

and permanent inactivation (Lee and Kaletunc, 2002; Mackey

demand for heating energy (kWh) per wet ton of treatedoval calculated for each pathogen indicator in that reactorconfigurations at 60 and 70 �C are based on conservativented here as upper bounds.

Heating Energy Demand to Achieve 1 log ExpectedInactivation (kWh/wet metric ton treated)

E. coli E. faecalis MS-2

24 17 39

15 16 19

14 14 16

11 10. 16

10 8.7 11

< 0.1 < 0.1 2.2

< 0.0001 < 0.0001 1.7

Treatment Scheme

MAD TAD 50 TAD 55 TPAD 50 TPAD 55 60+MAD 70+MAD

En

erg

y C

osts an

d B

en

efits

(k

Wh

p

er m

etric

to

n tre

ate

d s

lu

dg

e)

-50

0

50

100

150

200

250

300

Heat-up Cost

Heat Loss Cost

Biogas Value

Net Energy

Fig. 4 e Energy required and production per wet metric ton of treated sludge for various treatment schemes. Heat-up

requirement refers to the initial heating to bring the sludge up to temperature in the first or only phase of the digester. Heat

loss refers to heat energy lost from the digester or digester phases which must be replaced. Biogas production is expressed

as heat energy produced and reflects losses for combustion efficiency and moisture removal.

wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 4 7 5 8e4 7 6 8 4765

et al., 1991, 1993). Lower inactivation rates and documented

full-scale complications with VBNC behavior at 50 or 55 �Care likely caused by the absence of a true bacterial inactiva-

tion mechanism. The differences between the culturability

loss and viability loss (Fig. 3) at temperatures below 60 �Csuggests that E. coli treated in this temperature range may be

more prone to VBNC and reactivation behavior previously

observed in full-scale, thermophilic digesters. Loss of viability

was clearly below the maximum observable loss (3.7 log) for

50 (0.76 log) and 55 �C (1.31 log) temperatures. At 60 �C, theexposure time which causes a 4 log reduction in E. coli cul-

turability corresponds to a 2.14 log loss in viability, which is

also significantly less than the 3.7 log reduction in the auto-

claved control. The threshold for permanent ribosomal

damage in E. coli is approximately 60 �C. It is possible that the

variation in this value, or variation in experiment placed the

bacteria just below this lethal threshold. Experiments at 70 �Care more certainly lethal in terms of permanent ribosomal

damage, but due to the very rapid inactivation (4 log loss of

culturability in 16 s) the higher variation between replicates

at this temperature can likely be attributed to experimental

imprecision. The useful range of CTC staining is limited by

minimum detection limits of microscopy and non-specific

interactions between the stain and high concentrations of

inactivated cells. The maximum CTC-derived log reduction in

viability observed was 3.7 log in an autoclaved negative

control. While further investigation using different types of

organisms is necessary to confirm these effects, this work

does suggest that digestion or treatment above 60 �C provides

some advantage for reducing potential coliform reactivation.

Particular attention should be paid to exactly where the

threshold of permanent ribosomal damage falls for bacteria

of concern. Finally, the sharp inactivation rate increase in

bacteria at 60 �C was not observed in MS-2 phage, which does

not contain ribosomes. The greater resistance to thermal

inactivation also supports the use of MS-2 or coliphages as

a more conservative pathogen inactivation indicator than

fecal coliforms. Additionally, the inactivation rate coeffi-

cients observed for MS-2 are either similar or more resistant

to heat inactivation than the rate coefficients reported for

Ascaris suum (Aitken et al., 2005; Popat et al., 2010).

The relationship between inactivation rate constant and

digester temperature has also been investigated with Arrhe-

nius plots for each indicator organism, and is presented in the

Supplementary Data Figure S-4. The Arrhenius plot for MS-2

displays linearity through the entire temperature range

(25e70 �C). This indicates that the activation energy barrier for

MS-2 inactivation is constant through this temperature range,

and dominated by a single mechanism, likely protein dena-

turation. Arrhenius plots for E. coli and E. faecalis however,

show a linear region at lower temperatures and then amarked

increase at temperatures beyond 55 �C. This behavior

supports the conclusion that a threshold temperature exists

in bacteria whereby vital proteins or ribosomes become

completely and irreversibly denatured. For the important

indicators considered here, this temperature is above 55 �C.The shape of these Arrhenius plots is also consistent with the

markedly increased loss of bacteria viability observed above

60 �C (Fig. 3).

The log inactivation projections presented here for

common thermophilic and mesophilic temperatures are

consistent with previous bench and full-scale inactivation

observations. First-order inactivation rate coefficients for E.

coli, E. faecalis and MS-2 are similar to previously reported

values in several studies, conducted in water or manure, at

temperatures ranging from 35 to 60 �C (Aitken et al., 2007;

Nappier et al., 2006; Olsen and Larsen, 1987; Spinks et al.,

2006). Additionally, observations in full-scale sludge digester

systems for fecal coliforms have similarly demonstrated 1 to 2

log reductions in MAD, 3-4 log in TAD or TPAD and compete

inactivation by pasteurization at 70 �C (Bagge et al., 2005;

Sidhu and Toze, 2009; Viau and Peccia, 2009b).

4.2. Comparing energy efficiencies between digesterconfigurations

An energy balance for each digester configuration is domi-

nated by the value of biogas produced. This assessment

wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 4 7 5 8e4 7 6 84766

agrees with previous modeling studies that have suggested

net energy gains can be achieved by adding 70 �C pretreat-

ment steps at 1e5 day residence times (Bolzonella et al., 2007;

Lu et al., 2008). Energy requirements for heating consume

between 19 and 32% of this biogas value depending on the

reactor configuration. While no optimized energy balance

comparing thermophilic and mesophilic digestion conditions

has been published, our estimate of MAD heating require-

ments consuming 21% of biogas energy production (30% if

losses associated with heat recovery are neglected) compares

well with the previously published 35% for a MAD system

which neglects the potential for process heat recovery

(Lubken et al., 2007). These similarities and the fact that the

heating requirement and biogas models are well accepted for

estimating digester performance suggest that our model

provided heating and biogas production in the ranges that

are observed in full-scale practice and is useful for making

comparisons between digester configurations.

All 7 digestion configurations analyzed produce compa-

rable amounts of net energy per unit of treated sludge,

which indicates that 60 or 70 �C pretreatment can be

implemented without forfeiting energy production. While

this result is not intuitive, there are two reasons for why

high temperature digestion can be achieved without sacri-

ficing energy output. First, the heating costs in a reactor are

a combination of the heat input to bring the sludge up to the

correct temperature and the additional heat needed to

compensate for heat losses to the environment. Increasing

the reactor temperature increases the required initial heat

input as well as the rate of heat loss. However this effect is

mitigated by the dramatic reduction in residence times and

reactors sizes used at higher temperatures due to more

rapid inactivation. Secondly, our model incorporates

increases in biogas energy yield (based on 7% increase in

total VS destruction over MAD alone) for configurations

incorporating 50, 55, 60 or 70 �C treatment. Operating single

phase digesters at thermophilic temperatures or the pres-

ence of an initial thermophilic phase typically improves

hydrolysis and increases the bioavailability of digestible

material (Bolzonella et al., 2007; Ferrer et al., 2008; Lu et al.,

2008). A representative increase of 7% in total VS conversion

for thermophilic TPAD treatment has been noted in full-

scale digesters relative to MAD treatment alone (Shimp

et al., 2003). Slightly higher 9 and 10% increases have been

observed in bench-scale testing for 90 min of 60 �Cpretreatment (Salsabil et al., 2010) and short residence time

(2 day) TPAD systems (Ge et al., 2010) respectively, each

relative to MAD treatment alone.

Though not knowing the specific increase in biogas

production due to each pretreatment regime introduces some

model uncertainty, the impact on this uncertainty is minimal.

If our model did not include an increase in biogas production

for higher temperature treatments, the decrease in net energy

associated with adding pretreatment to MAD would only be

a loss of 1% at 70 �C or 3% at 60 �C, which does not impact our

broader conclusions about relative net energy production

between thermophilic and mesophilic processes. Net energy

calculations for all reactor configurations are also affected by

environmental assumptions, such as air, soil, and initial

sludge temperatures. However as the net energy calculations

shift across different environmental temperature ranges, the

relative difference in expected net energy between our reactor

configurations of interest only differs by 1 or 2 percent from

the relationship expressed in Fig. 4.

5. Conclusion

Traditionally, anaerobic digestion has been designed and

optimized with the goals of increased gas production and

solids destruction. Optimizing for the inactivation of patho-

gens has been given little attention. The results presented here

describe the potential for significantly greater inactivation of

bacterial and viral pathogens in biosolids and demonstrate

that it is possible to do so while working within currently

available configurations andwithout increases in energy costs.

� The temperature-based inactivation rates observed in batch

reactors demonstrate a dramatic increase in pathogen

destruction potential when a 60 or 70 �C phase is included.

At 60 �C, inactivation rates doubled in MS-2 phage and

increased 2 orders of magnitude for enteric bacteria surro-

gates over rates for the more common 50 to 55 �C thermo-

philic conditions. More dramatic increases in inactivation

rates were achieved at 70 �C. Thermal treatment of E. coli to

produce w4 logs of culturable inactivation demonstrated

that at temperatures less than 70 �C, the loss of cell viability

was not commensurate to the loss of culturability.

� Although significantly larger reductions in pathogen indi-

cators are observed in the high temperature regimes, the net

energy production in the 60 or 70 �C pretreatment systems

was not significantly different than the net energy produced

in more conventional systems that operate in mesophilic

(w37 �C) and thermophilic (50e55 �C) ranges.

Appendix. Supplementary material

Supplementary data related to this article can be found online

at doi:10.1016/j.jorganchem.2011.03.010.

r e f e r e n c e s

Aitken, M.D., Sobsey, M.D., Blauth, K.E., Shehee, M., Crunk, P.L.,Walters, G.W., 2005. Inactivation ofAscaris suum and poliovirusinbiosolids under thermophilic anaerobic digestion conditions.Environmental Science & Technology 39 (15), 5804e5809.

Aitken, M.D., Sobsey, M.D., Van Abel, N.A., Blauth, K.E.,Singleton, D.R., Crunk, P.L., Nichols, C., Walters, G.W.,Schneider, M., 2007. Inactivation of Escherichia coli O157:H7during thermophilic anaerobic digestion of manure from dairycattle. Water Research 41 (8), 1659e1666.

APHA, AWWA, WEF, 2005. Standard Methods for the Examinationof Water and Wastewater. D.C, Washington.

Bagge, E., Sahlstrom, L., Albihn, A., 2005. The effect of hygienictreatment on the microbial flora of biowaste at biogas plants.Water Research 39 (20), 4879e4886.

Beecher, N., Crawford, K., Goldstein, N., Kester, G., Lono-Batura, M., Dziezyk, E., 2007. A National Biosolids Regulation,

wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 4 7 5 8e4 7 6 8 4767

Quality, End Use & Disposal Survey. North East Biosolids andResiduals Association.

Bekkering, J., Broekhuis, A.A., van Gemert, W.J.T., 2010.Optimisation of a green gas supply chain - a review.Bioresource Technology 101 (2), 450e456.

Bolzonella, D., Pavan, P., Zanette, M., Cecchi, F., 2007. Two-phaseanaerobic digestion of waste activated sludge: effect of anextreme thermophilic prefermentation. Industrial andEngineering Chemistry Research 46 (21), 6650e6655.

Evanylo, G., Sherony, C., Spargo, J., Starner, D., Brosius, M.,Haering, K., 2008. Soil and water environmental effects offertilizer-, manure-, and compost-based fertility practices inan organic vegetable cropping system. Agriculture,Ecosystems & Environment 127 (1e2), 50e58.

Ferrer, I., Ponsa, S., Vazquez, F., Font, X., 2008. Increasing biogasproduction by thermal (70 �C) sludge pre-treatment prior tothermophilic anaerobic digestion. Biochemical EngineeringJournal 42 (2), 186e192.

Funderburg, S.W., Sorber, C.A., 1985. Coliphages as indicators ofenteric viruses in activated sludge. Water Research 19 (5),547e555.

Gantzer, C., Gaspard, P., Galvez, L., Huyard, A., Dumouthier, N.,Schwartzbrod, J., 2001. Monitoring of bacterial andparasitological contamination during various treatment ofsludge. Water Research 35 (16), 3763e3770.

Ge, H., Jensen, P.D., Batstone, D.J., 2010. Pre-treatmentmechanisms during thermophilic-mesophilic temperaturephased anaerobic digestion of primary sludge. Water Research44 (1), 123e130.

Greer, D., 2007. Financing an anaerobic digester. BioCycle 48 (12),44e48.

Guzman, C., Jofre, J., Montemayor, M., Lucena, F., 2007.Occurrence and levels of indicators and selected pathogens indifferent sludges and biosolids. Journal of AppliedMicrobiology 103 (6), 2420e2429.

Havelaar, A.H., Van Olphen, M., Drost, Y.C., 1993. F-specific RNAbacteriophages are adequate model organisms for entericviruses in fresh water. Applied and EnvironmentalMicrobiology 59 (9), 2956e2962.

Higgins, M.J., Chen, Y.C., Murthy, S.N., Hendrickson, D., Farrel, J.,Schafer, P., 2007. Reactivation and growth of non-culturableindicator bacteria in anaerobically digested biosolids aftercentrifuge dewatering. Water Research 41 (3), 665e673.

Hobbie, J.E., Daley, R.J., Jasper, S., 1977. Use of nuclepore filters forcounting bacteria by fluorescence microscopy. Applied andEnvironmental Microbiology 33 (5), 1225e1228.

Kepp, U., Machenbach, I., Weisz, N., Solheim, O.E., 2000.Enhanced stabilisation of sewage sludge through thermalhydrolysis three years of experience with full scale plant.Water Science & Technology, 89e96.

Khaleel, R., Reddy, K.R., Overcash, M.R., 1981. Changes in soilphysical properties due to organic waste applications:a review. Journal of Environmental Quality 10 (2), 133e141.

Krich, K., Augenstein, D., Batmale, J., Benemann, J., Rutledge, B.,Salour, D., 2005. Biomethane from Dairy Waste: a Sourcebookfor the Production and Use of Renewable Natural Gas inCalifornia (Prepared for Western United Dairymen).

Le, M.S., Mayhew, M.E., Back, P.A., 2002. Effectiveness ofsecondary digesters as a pathogen controller in winter. Waterand Environment Journal 16 (4), 292e295.

Lee, J., Kaletunc, G., 2002. Evaluation of the heat inactivation ofEscherichia coli and Lactobacillus plantarum by differentialscanning calorimetry. Applied and EnvironmentalMicrobiology 68 (11), 5379e5386.

Lewis, D., Gattie, D.K., 2002. Pathogen risks from applying sewagesludge to land. Environmental Science and Technology 36 (13),286Ae293A.

Low, S.-Y., Baertsch, C., Paez-Rubio, T., Kucharski, M., Peccia, J.,2007. Off-site exposure to respirable aerosols produced during

the disk incorporation of class B biosolids. Journal ofEnvironmental Engineering 133, 897e994.

Lu, J., Gavala, H.N., Skiadas, I.V., Mladenovska, Z., Ahring, B.K.,2008. Improving anaerobic sewage sludge digestion byimplementation of a hyper-thermophilic prehydrolysisstep. Journal of Environmental Management 88 (4),881e889.

Lubken, M., Wichern, M., Schlattmann, M., Gronauer, A., Horn, H.,2007. Modelling the energy balance of an anaerobic digesterfed with cattle manure and renewable energy crops. WaterResearch 41 (18), 4085e4096.

Mackey, B.M., Miles, C.A., Parsons, S.E., Seymour, D.A., 1991.Thermal-denaturation of whole cells and cell components ofEscherichia coli examined by differential scanning calorimetry.Journal of General Microbiology 137, 2361e2374.

Mackey, B.M., Miles, C.A., Seymour, D.A., Parsons, S.E., 1993.Thermal denaturation and loss of viability in Escherichia coliand Bacillus stearothermophilus. Letters in Applied Microbiology16 (2), 56e58.

Metcalf and Eddy, 2003. Wastewater Engineering: Treatment andReuse. Mc Graw-Hill, New York.

Nappier, S.P., Aitken, M.D., Sobsey, M.D., 2006. Male-specificcoliphages as indicators of thermal inactivation of pathogensin biosolids. Applied and Environmental Microbiology 72 (4),2471e2475.

NRC, 2002. Biosolids Applied to Land: Advancing Standards andPractices. National Research Council of the NationalAcademies, Washington D.C.

Olsen, J.E., Larsen, H.E., 1987. Bacterial decimation times inanaerobic digestions of animal slurries. Biological Wastes 21(3), 153e168.

Pedersen, D.C., 1981. Density Levels of Pathogenic Organisms inMunicipal Wastewater Sludge: a Literature Review. U.S. EPAMunicipal Environmental Research Laboratory, Cincinnati,OH. EPA 600-2-81-170.

Perry, R.H., Green, D.W., 1984. Perry’s Chemical Engineers’Handbook. McGraw-Hill Book Co., New York.

Popat, S.C., Yates, M.V., Deshusses, M.A., 2010. Kinetics ofinactivation of indicator pathogens during thermophilicanaerobic digestion. Water Research 44 (20), 5965e5972.

Qi, Y., Dentel, S.K., Herson, D.S., 2007. Increases in fecal coliformbacteria resulting from centrifugal dewatering of digestedbiosolids. Water Research 41 (3), 571e580.

Rittmann, B.E., McCarty, P.L., 2001. Environmental Biotechnology:Principles and Applications. McGraw-Hill, New York.

Sahlstrom, L., Aspan, A., Bagge, E., Danielsson-Tham, M.-L.,Albihn, A., 2004. Bacterial pathogen incidences in sludge fromSwedish sewage treatment plants. Water Research 38,1989e1994.

Salsabil, M.R., Laurent, J., Casellas, M., Dagot, C., 2010. Techno-economic evaluation of thermal treatment, ozonation andsonication for the reduction of wastewater biomass volumebefore aerobic or anaerobic digestion. Journal of HazardousMaterials 174 (1e3), 323e333.

Shimp, G.F., Rowan, J.M., Long, D.W., Santha, H., 2003. Anaerobicdigestion. retooling an old process to meet a ‘Class A’objective. Water Environment and Technology 15 (5), 45e49.

Sidhu, J.P.S., Toze, S.G., 2009. Human pathogens and theirindicators in biosolids: a literature review. EnvironmentInternational 35 (1), 187e201.

Spinks, A.T., Dunstan, R.H., Harrison, T., Coombes, P., Kuczera, G.,2006. Thermal inactivation of water-borne pathogenic andindicator bacteria at sub-boiling temperatures. WaterResearch 40 (6), 1326e1332.

USEPA, 1979. Process Design Manual Sludge Treatment andDisposal. U.S. EPA Office of Research and Development,Washington, D.C. EPA 625-1-79-011.

USEPA, 1994. Land Aplication of Sewage Sludge. A Guide for LandAppliers on the Requirements of the Federal Standard for the

wat e r r e s e a r c h 4 5 ( 2 0 1 1 ) 4 7 5 8e4 7 6 84768

Use or Disposal of Sweage Sludge 40 CFR Part 503, U.S. EPAOffice of Enforcement and Compliance Assurance,Washington, D.C EPA 831-B-93002b.

USEPA, 1999. Environmental Regulations and Technology: Controlof Pathogens and Vector Attraction in Sewage Sludge. U.S. EPAOffice of Research and Development, Washington, D.C. EPA625-R-92013.

USEPA, 2001. Method 1601: Male-specific (Fþ) and SomaticColiphage in Water by Two-Step Enrichment Proceedure. U.S.EPA Office of Water, Washington, D.C. EPA 821-R-01e030.

USEPA, 2002. Method 1600: Enterococci in Water by MembraneFiltration Using Membrane-Enterococcus Indoxyl-b-D-

Glucoside Agar (MEI). U.S. EPA Office of Water, Washington,D.C. EPA 821-R-02e022.

Viau, E., Peccia, J., 2009a. Evaluation of the enterococci indicatorin biosolids using culture-based and quantitative PCR assays.Water Research 43 (19), 4878.

Viau, E., Peccia, J., 2009b. A survey of wastewater indicators andhuman pathogen genomes in biosolids produced by class Aand class B stabilization treatments. Applied andEnviornmental Microbiology 75, 164e174.

Zupancic, G.D., Ros, M., 2003. Heat and energy requirements inthermophilic anaerobic sludge digestion. Renewable Energy 28(14), 2255e2267.