Tese de doutorado - REPOSITORIO INSTITUCIONAL DA UFOP

Transcription

Tese de doutorado
“ESTUDO DA REDUÇÃO DE SULFATO EM
REATORES CONTÍNUOS UTILIZANDO
GILCEROL”
Autora:
Orientador:
Co-orientador:
Sueli Moura Bertolino
Prof. Versiane Albis Leão
Prof.Sérgio Francisco de Aquino
MAIO DE 2012
Sueli Moura Bertolino
“ESTUDO DA REDUÇÃO DE SULFATO EM REATORES
CONTÍNUOS UTILIZANDO GLICEROL”
Tese de Doutorado apresentada ao Programade
Pós-Graduação em Engenharia de Materiais da
REDEMAT, como parte integrante dos requisitos
para a obtenção do título de Doutor em
Engenharia de Materiais.
Área de concentração: Processos de fabricação
Orientador: Prof. Versiane Albis Leão
Coorientador: Prof. Sérgio Francisco de Aquino
Ouro Preto, maio de 2012
B546e
Bertolino, Sueli Moura.
Estudo da redução de sulfato em reatores contínuos utilizando glicerol
[manuscrito] / Sueli Moura Bertolino. – 2012.
xii, 138 f.: il. color.; grafs.; tabs.
Orientador: Prof. Dr. Versiane Albis Leão.
Coorientador: Prof. Dr. Sérgio Francisco de Aquino.
Tese (Doutorado) - Universidade Federal de Ouro Preto. Escola
de Minas. Rede Temática em Engenharia de Materiais.
Área de concentração: Processos de Fabricação.
1. Bactérias redutoras de sulfato - Teses. 2. Reatores anaeróbios - Reator
UASB - Teses. 3. Glicerol - Teses. 4. Biotecnologia - Teses. 5. Fluidização Teses. I. Universidade Federal de Ouro Preto. II. Título.
Catalogação: sisbin@sisbin.ufop.br
i
ii
"A ciência humana de maneira nenhuma nega a existência de Deus.
Quando considero quantas e quão maravilhosas coisas o homem
compreende, pesquisa e consegue realizar, então reconheço claramente
que o espírito humano é obra de Deus, e a mais notável."
Galileu Galilei (Fé e ciências)
iii
“Dedico esta tese de doutorado a duas incríveis pessoas: a meu pai, Milton Bertolino,
que além me fazer a pessoa que sou, me deixou o que eu tenho de mais valor: a vontade
de viver. A Terezinha Bertolino, minha mãe, uma professora, uma amiga, uma mulher
inigualável que me ensinou a como viver”
iv
AGRADECIMENTOS
À Deus, pelo dom da vida.
Obrigada mãe, por ser meu exemplo de vida! Uma professora que orgulha de ter como maior
recompensa, o reconhecimento de tantos alunos, hoje adultos, que tiveram a honra de ser
alfabetizados pela senhora. Uma mãe, que mesmo cansada de sua jornada de trabalho, tinha
sempre um lindo sorriso para nos receber. Uma mulher, que tira em Deus toda sua fé para
superar as dificuldades e nos mostrar que a vida é nosso maior presente de Deus. Te amo
mãe!
À minha família, que é o alicerce da minha vida. Agradeço o amor e o companheirismo de
meus irmãos, Rosana, Mílber, Simone e Nélio. A meu cunhado Zé, e minhas cunhadinhas Lu
e Fabi. Aos meus lindos sobrinhos, Helena, Thiago, Pedro, Ana Clara e Ana Laura.
À todos os meus familiares,“Mouras e Bertolinos”, que fazem minha vida fazer sentido.
Ao meu orientador.É indiscutível seu comprometimento com a pesquisa e orientação de seus
alunos, o que é refletido no nosso aprendizado.Foram muitas reuniões, muitas conversas,
muitos erros e acertos, até chegar aqui. Muito obrigada,professor Versiane, pela confiança e
pela enorme contribuição na minha formação.
Ao professor Sérgio, pela co-orientação através das correções e sugestões durante a execução
desta tese de doutorado, principalmente, na elaboração dos artigos.
v
À equipe do laboratório de Biologia Molecular da UFOP, principalmente, professora Renata e
a mestranda Isabel.
Às eternas amigas/irmãs, unidas desde: o pré-escolar (Érica e Quel), o ensino fundamental
(Fábia e Cris), o ensino médio (Bi e Juli), a graduação (Fafá, Mi, Angel e San), o mestrado
(Jojô) e, também,à amiga Cintia. Obrigada amigas, vocês entraram em momentos diferentes
na minha vida, mas todasfazem parte da minha história.
À família SELETA! Jojo, Wandinha, Bia, Lili, Lessandra, Na, Mandinha, Nininha, Kkau, e
aos agregados, Emerson, Alan e Chatubinha. Obrigada seleteiras e seleteiros pela amizade e
companheirismo. Em particular, às seleteiras, agradeço pelas risadas, pelos bate-papos, pelas
conversas sérias, pela paciência em me
aturar nos momentos de cansaço e,
principalmente,pela oportunidade de ensinar e também de aprender. Amo todas vocês!
Aos alunos de IC, Frederico, Jamily, Nayara e Lucas, que com muito comprometimento me
ajudaram a cuidar de meus reatores e de minhas bactérias.
Ao técnico de laboratório, Sérgio, por todas as análises realizadas para a execução deste
projeto e, também, pela troca de conhecimentos.
A todos os colegas do laboratório de bio-hidrometalurgia, em especial minhas colegas de sala
Damaris e Larissa, por toda ajuda e companheirismo.
À Vale, pelo financiamento que possibilitou o desenvolvimento técnico/científico desta tese
de doutorado.
Ao Conselho Nacional de Pesquisa pelo financiamento da bolsa de doutorado.
À UFOP e a REDEMAT, pela importante participação em minha formação.
vi
RESUMO
A hidrodinâmica do reator desempenha um papel chave durante a redução do sulfato porque
bactérias redutoras de sulfato (SRB) não formam lodo granular facilmente. Além disso, entre
os maiores desafios para a implementação da bioredução do sulfato estão o custo da fonte de
elétrons e sua disponibilidade.Nesta tese, o desempenho de um reator anaeróbio de fluxo
ascendente e manta de lodo (UASB) foi comparado com o de um reator de leito fluidizado
(RLF), tratando efluente sintético contendo lactato como fonte de carbono e de elétrons.A
carga orgânica, a redução do sulfato e as condições de mistura foram os principais parâmetros
monitorados. O perfil dos ácidos graxos voláteis (AGV) e técnicas moleculares permitiram
propor as vias metabólicas envolvidas durante a degradação do lactato. Para altas cargas
orgânicas observou-se que: (i) olactato foi oxidado a acetato e dióxido de carbono por
bactérias que oxidam incompletamente o substrato (Desulfomonas, Desulfovibrio,
Desulfolobus, Desulfobulbus e Desulfotomaculum spp.); (ii) o lactato foi convertido a acetato
por bactérias fermentativas (BF), tais como Clostridium sp. Sem recirculação, o reactor
UASB apresentou uma taxa de redução volumétrica máxima de sulfato de 1,3g/(L.d)) (66%
de remoção), enquanto, concentrações elevadas de propionato, no efluente, estavam
associadas abaixas eficiências de redução de sulfato, um resultado da competição entre as BF
e BRS pelo substrato. A recirculação da biomassa melhorou consideravelmente a eficiênica de
redução do sulfato para 89% (taxa específica de 0,089±0.014g/(gSSV.d)), para uma razão
DQO/sulfato de 2,5±0,2.No entanto, valores duas vezes mais elevados (0,191 ± 0.016g /
(gSSV.d)) foram obtidos no RLF, tratando o mesmo substrato.Nas melhores condições
operacionais, o RLF apresentou uma eficiência de redução de sulfato de 97% (64mg/L de
sulfato residual) e a atividade fermentativa foi desprezível durante a degradação do lactato. O
RLF foi então selecionado para avaliar o glicerol como uma fonte de carbono alternativa e o
desempenho da redução do sulfato foi comparado com o obtido durante a degradação do
lactato. A redução do sulfato na presença de glicerol produziu uma DQO residual (1700mg/L)
menor do que a produzida com lactato (2500mg/L) para a mesma razão DQO/sulfato
(2.5).Estimou-se que 50% da degradação do glicerol foi devida a redução de sulfato e 50%
àfermentação, o que foi confirmadopela presença de butirato no efluente do RLF.O reator
UASB foi incapaz de produzir uma concentração de sulfatoabaixo de 250mg/L, devido às
condições inadequadas de mistura. Por outro lado, o RLF efetivamente produziu um efluente
com concentrações de sulfato abaixo do valor referência. O glicerol pode ser uma alternativa
de baixo custo eficaz para a redução do sulfato e esta biotecnologia mais uma aplicação para o
tratamento dos resíduos gerados na indústria de biodiesel.
v
ABSTRAT
Reactor hydrodynamics plays a key role during sulfate reduction because sulfate reducing
bacteria (SRB) do not granulate easily. In addition, one of the greatest challenges for the full
implementation of biological sulfate reduction is the cost of the electron source along with its
availability. In this thesis, the performance of an Upflow Anaerobic Sludge Blanket (UASB)
reactor was compared to that of a fluidized bed reactor (FBR), treating lactate as carbon and
electron source. Organic loading, sulfate reduction and mixing conditions were the main
parameters monitored. Volatile fatty acids (VFA) profile and molecular biology techniques
enabled the assessment of the metabolic pathways accounting for lactate degradation. At high
organic loadings, it was observed that: (i) lactate was oxidized to acetate and carbon dioxide
by incomplete-oxidizing SRB (Desulfomonas, Desulfovibrio, Desulfolobus, Desulfobulbus
and Desulfotomaculum spp.); (ii) lactate was converted to acetate by fermenting bacteria (FB)
such as Clostridium sp. Without recirculation, the UASB reactor showed a maximum
volumetric sulfate reduction rate of 1.3g/(L.d)) (66% removal), while high propionate
concentrations were associated to low sulfate reduction efficiencies, a result of the
competition between FB and SRB for the substrate. Biomass recirculation considerably
improved sulfate reduction yields to 0.089±0.014g/(gSSV.d), (89% reduction), for an
COD/sulfate mass ratio value of 2.5±0.2. Nevertheless, values twice as higher
(0.191±0.016g/(gSSV.d)) were achieved in the FBR, treating the same substrate. In the best
operational conditions, the FBR depicted a sulfate reduction efficiency of 97% (64mg/L
residual sulfate) and negligible fermentative activity during lactate oxidation. It was then
selected for experiments utilizing glycerol as an alternative carbon source and the sulfate
reduction performance was compared to that observed with lactate. Sulfate reduction in the
presence of glycerol produced residual COD (1700mg/L) smaller than that produced with
lactate (2500mg/L C2 H3O2) at the same COD/sulfate mass ratio (2.5). It was estimated that
50% of glycerol degradation was due to sulfate reduction and 50% to fermentation, which
was supported by an increased butyrate concentration in the FBR effluent. The UASB reactor
was unable to produce final sulfate concentrations below 250mg/L due to poor mixing
conditions. Conversely, the FBR consistently ensured residual sulfate concentrations below
the target value. Glycerol can be a cost-effective alternative for sulfate reduction and a viable
solution and this biotechnology a new application for residues generated in the biodiesel
industry.
vi
SUMÁRIO
LISTA DE FIGURAS .......................................................................................................... ix
LISTA DE TABELAS.......................................................................................................... xi
LISTA DE NOTAÇÕES ..................................................................................................... xii
CAPÍTULO 1 .........................................................................................................................1
1.1 Introdução.....................................................................................................................1
1.2. Objetivos e organização da tese..................................................................................13
1.3. Referências ................................................................................................................16
CAPÍTULO 2 ....................................................................................................................... 22
PERFORMANCE OF CONTINUOUS BIOREACTORS FOR SULFATE REDUCTION
AIMING AT UTILIZING GLYCEROL AS CARBON SOURCE ........................................ 22
2.1. Introduction ...............................................................................................................23
2.2. Materials and methods ...............................................................................................25
2.2.1. Bioreactors ..........................................................................................................25
2.2.2. Microorganisms and reactor start-up ....................................................................26
2.2.3. Analytical methods ..............................................................................................28
2.3. Results .......................................................................................................................29
2.3.1. Performance and stability of UASB and Fluidized Bed reactors. ..........................29
2.3.2. COD consumption and sulfate reduction yields ....................................................35
2.3.3. Effect of the reactor configuration in sulfate removal ...........................................36
2.3.4. Sulfate reduction in the presence of pure glycerol as substrate .............................41
2.4. Discussion .................................................................................................................42
2.5. Conclusions ...............................................................................................................47
2.6. Acknowledgements ....................................................................................................48
2.7. References .................................................................................................................48
CAPÍTULO 3 ....................................................................................................................... 54
IMPLICATIONS OF VOLATILE FATTY ACID PROFILE ON THE METABOLIC
PATHWAY DURING CONTINUOUS SULFATE REDUCTION ....................................... 54
3.1. Introduction ...............................................................................................................55
3.2. Experimental..............................................................................................................57
3.2.1. Microorganisms and growth medium ...................................................................57
3.2.2. Anaerobic reactor and operational methods..........................................................57
vii
3.3. Results and discussion ...............................................................................................60
3.3.1. Reactor start-up and biomass ...............................................................................60
3.3.2. Reactor performance ............................................................................................63
3.3.3. Influence of effluent recirculation in sulfate reduction .........................................76
3.4. Conclusions ...............................................................................................................77
3.5. Acknowledgements ....................................................................................................78
3.6. References ................................................................................................................78
CAPÍTULO 4 ....................................................................................................................... 85
GLYCEROL AS AN ELECTRON DONOR FOR SULFATE REDUCTION IN FLUIDIZED
BED REACTORS ................................................................................................................ 85
4.1. Introduction ...............................................................................................................86
4.2. Experimental..............................................................................................................88
4.2.1. Anaerobic reactor.................................................................................................88
4.2.2. Microorganisms and reactor start-up ....................................................................89
4.2.3. Operational methods ............................................................................................90
4.2.4. Batch experiments with glycerol ..........................................................................91
4.3. Results and discussion ...............................................................................................93
4.3.1. Reactor start-up and biomass ...............................................................................93
4.3.2. Reactor performance ............................................................................................95
4.4. Conclusions ............................................................................................................. 108
4.5. Acknowledgements ................................................................................................... 109
4.6. References ............................................................................................................... 109
CAPÍTULO 5 ..................................................................................................................... 112
CAPÍTULO 7 ..................................................................................................................... 117
ANEXOS ........................................................................................................................... 118
viii
LISTA DE FIGURAS
Figure 2.1 Pictures of the two lab-scale reactors, UASB and FBR. Port c in the UASB reactor
was utilized for biomass recirculation during phase VII. In the FBR biomass was performed
from point g. .........................................................................................................................25
Figure 2.2.Time diagram showing experimental conditions applied in both the UASB reactor
and the FBR. Inside each box is depicted the COD/sulfate mass ratio. When there was a
change on the COD or sulfate loading the other parameter was kept constant. During phase VI
(UASB reactor), the change on the COD/Sulfate ratio was due to different flow rate applied.
.............................................................................................................................................27
Figure 2.3. Sulfate removal, residual and target sulfate concentrations in different phases of
the FBR (A) and UASB reactor (B) operations. ....................................................................30
Figure 2.4. Performance parameters in different phases (according to time diagram, figure
2.2) in the FBR (A and C) and the UASB reactor (B and D).VFA: volatile fatty acids; BA:
bicarbonate alkalinity. ..........................................................................................................32
Figure 2.5.Volumetric organic and sulfate loading rates applied and removal in the UASB
reactor and the FBR. Organic loading and removal rates in the FBR (A) and UASB reactor
(B); sulfate loading and removal rates in the FBR (C) and UASB reactor (D). ......................33
Figure 2.6. SRB population and lactate oxidized by SRB during continuous sulfate removal
in the FBR (A) and UASB reactor (B).Glycerol was utilized as substrate in phase VI during
the operation of the FBR and therefore does not appear in figure A. .....................................34
Figure 2.7.Acetate, butyrate and propionate profiles in the FBR (a) and the UASB reactor
(b).Details on the different phases are depicted in figure 2.2. ................................................37
Figure 2.8.Biomass profile in the UASB reactor (ports A, B and C; figure 2.1) during phase
III (no recirculation) and VII (with recirculation). Port c during phase VII was utilized for
biomass recirculation, so VSS not was determinated.............................................................39
Figure 2.9. Values of specific sulfate-reduction and propionate production rates in the UASB
reactor. Phase VI is characterized by a change in both flow rate and lactate concentration. ...41
Figure 2.10. Main metabolic pathways developed during continuous sulfate removal in UASB
and FBR during lactate and glycerol degradation. FB - Fermenting Bacteria; SRB - Sulfate
Reducing Bacteria. ...............................................................................................................43
Figure 3.1. Schematic diagram of the UASB reactor for sulfate reduction. ...........................58
ix
Figure 3.2. Evolution of biomass monitored in the UASB reactor: (I) OLR = 3.48 kg/m3.d;
(II) OLR = 4.87 kg/m3.d; (III) OLR = 3.55 kg/m 3.d; (IV) OLR = 4.65 kg/m3.d; (V) OLR =
5.89 kg/m3.d; (VI) OLR = 5.04 kg/m3.d. ...............................................................................61
Figure 3.3. Performance parameters of the UASB reactor, in different phases (according to
table 3.1). VFA: volatile fatty acids; BA: bicarbonate alkalinity. ..........................................63
Figure 3.4. Parameters monitored during sulfate reduction with lactate. Volumetric organic
loading (A); COD removal efficiency (B). Volumetric sulfate loading (C). Sulfate removal
efficiency (D), during the phases I to VI (according to table 3.1). .........................................66
Figure 3.5. Metabolic pathways relevant in this study involving the anaerobic metabolism of
lactate. Species and reaction refer to identified microorganism (table 3.2) and anaerobic
degradation reactions (table 3.3). ..........................................................................................69
Figure 3.6. Comparison between estimated and analytical concentrations of: (A) acetate; (B)
propionate, during phases I to VI (table 3.1). OLR: organic load rate (kg/m3.d)....................70
Figure 3.7. Values of effluent- and CODVFA during sulfate reduction in the UASB reactor.
CODout was measured and CODVFA values were determined from the measured propionate
and acetate concentrations. ...................................................................................................75
Figure 4.1. Schematic diagram of the FBR reactor for sulfate reduction. ..............................89
Figure 4.2. Evolution of biomass concentration monitored on FBR reactor. .........................94
Figure 4.3. Parameters monitored during sulfate reduction with lactate and glycerol. COD
concentration (A) and removal efficiency (B). Sulfate concentration (C) and removal
efficiency (D). The first 50 days correspond to the adaptation period. The circle on figure
4.3D represents the period when only glycerol was fed to the FBR. ......................................97
Figure 4.4. Performance parameters of the FBR reactor, in different phases. The arrow in
figure 4.4A indicates a peak on VFA production................................................................. 100
Figure 4.5. Parameters monitored in the batch experiment.(A) bacterial growth (OD), sulfate
concentration and pH. (B) VFA profile and total organic carbon......................................... 106
Figure 4.6. Metabolic pathways for glycerol and lactate degradation during sulfate reduction.
X is an electron carrier........................................................................................................ 107
x
LISTA DE TABELAS
Tabela 1.1 Reações envolvidas no metabolismo de grupos de bactérias durante o processo
anaeróbio. ...............................................................................................................................4
Tabela 1.2. Desempenhos de vários reatores UASB usados para a redução de sulfato. ............9
Tabela 1.3.Desempenho de vários reatores de leito fluidizado usados para a redução do
sulfato. .................................................................................................................................11
Table 2.1.Characteristics and operating conditions of the reactors studied............................26
Table 2.2. Best parameters achieved during sulfate reduction in UASB reactor and the FBR,
treating synthetic sulfate wastewater with lactate (phases VII - UASB reactor and I – FBR) or
glycerol (phase VI - FBR).....................................................................................................40
Table 3.1. Operational parameters during sulfate reduction in the UASB reactor. Hydraulic
retention time: 24 hours, 25ºC...............................................................................................59
Table 3.2. Microorganisms identified by molecular biology techniques in the inoculum and
different phases during UASB reactor operation. ..................................................................62
Table 3.3. Anaerobic degradation reactions relevant to this study. ........................................64
Table 3.4.Parameters related to sulfate removal in the UASB reactor as a function of feed
lactate concentration. ............................................................................................................68
Table 4.1. Operational parameters during sulfate reduction in the FBR reactor. HRTa of 10
hours and 25ºC. ....................................................................................................................91
Table 4.2.Operational parameters during replacement of lactate by glycerol in the FBR reactor
with SLR of 3.62 ± 0.23Kg/m3.d, at HRT of 13 hours and 25ºC. ..........................................91
Table 4.3.Microorganisms identified by molecular biology techniques in the inoculum during
FBR operation. .....................................................................................................................95
Table 4.4. Anaerobic degradation reactions relevant to this study. ........................................99
Table 4.5.Parameters of incomplete lactate oxidation reaction as a function of feed lactate and
sulfate (according to reaction 4).Values in mmol/L. ............................................................ 101
Table 4.6. Parameters of incomplete glycerol and lactate oxidation reaction during the
beginning of phase V. Values in mmol/L. ........................................................................... 103
xi
LISTA DE NOTAÇÕES
AB
Alcalinidade Bicarbonato
AGV
Ácidos Graxos Voláteis
AnSBBR
“Anaerobic Sequencing Biofilm Batch Reactor”
BA
“Bicarbonate Alkalinity”
BF
Bactérias Fermentativas
BRS
Bactéria Redutora de Sulfato
COA
Carga Orgânica Aplicada
COD
“Chemical oxygen demand”
CONAMA
Conselho Nacional do Meio Ambiente
CSA
Carga de Sulfato Aplicada
CSR
Carga de Sulfato Removida
DAM
Drenagem Ácida de Mina
DQO
Demanda Química de Oxigênio
EGSB
“Expanded Granular Sludge Blanket”
ETE
Estação de Tratamento de Esgoto
FB
“Fermentatives bactérias”
FBR
“Fluidized bed reactor
HPLC
“High Performance Liquid Chromatography”
HRT
“Hydraulic retention time”
INAP
“International Network for Acid Prevention”
MPM
“Micoorganisms producing methane”
ORL
“Organic Rate Loading”
RAHLF
Reator Anaeróbio Horizontal de Leito Fixo
xii
RLF
Reator de leito fuidizado
SLR
“Sulfate Loading Rate”
SSV
Sólidos Suspensos Voláteis
TDH
Tempo de Detenção Hirdráulica
TDS
“Total Dissolved Solids”
TOC
“Total organic carbon”
UASB
“Upflow Anaerobic Sludge Blanket”
UFOP
Universidade Federal de Ouro Preto
VFA
“Volatile Fatty Acids”
VSS
“Volatile suspended solids”
WHO
“World Health Organization”
1,3-PD
“1,3-propanediol”
xiii
CAPÍTULO 1
1.1 Introdução
A geração da drenagem ácida de mina (DAM) pode ser considerada como um dos impactos
responsáveis pela regulação de efluentes com altas concentrações de sulfato, o que levou à
crescente busca de tecnologias para o seu tratamento. A DAM é formada quando minerais
sulfetados presentes em resíduos da mineração (rejeito ou estéril) são oxidados em presença
de água, liberando uma solução ácida. Como consequência, no meio hídrico, ocorre o
aumento da acidez e da concentração de metais tóxicos, bem como da concentração de
sulfato(AKCIL e KOLDAS, 2006). O sulfato está também presente em efluentes de outras
indústrias tais como, a de papel, curtumes,a de alimentos, a de explosivos, a de tensoativos, a
de xenobióticos e aquímica/metalúrgica. Nestes efluentes industriais, a concentração de
sulfato pode chegar a 8g/L(LENSet al., 1995; WHO, 2011).
A presença do íon sulfato em efluentes industriais está associada aos possíveis efeitos tóxicos
causados pelo íon sulfeto que se forma no processo anaeróbio da redução de sulfato. Este
último pode comprometer a qualidade dos corpos d´água devido ao aumento da demanda
química de oxigênio (LENSet al., 1998). Além disso, o reuso de tais efluentes pelas indústrias
é, normalmente, um processo inviável, pois a presença de sulfeto promove a corrosão de
tubulações, estruturas e equipamentos, tornando-se, portanto, necessário o desenvolvimento
de técnicas eficientes e de baixo custo que possam ser utilizadas no tratamento de efluentes
com tais características (INAP, 2003; WHO, 2011).
De acordo com a Organização Mundial de Saúde (WHO, 2011), ainda não se têm dados que
permitam afirmar qual o nível de sulfato capaz de causar efeitos adversos aos seres humanos.
1
Pessoas adultas que consomem água potável contendo sulfato em concentrações superiores a
600 mg/L podem apresentar quadro de diarréia, embora também seja reportado que, com o
tempo, os humanos podem se adaptar a altas concentrações do íon(WHO, 2011). Em função
dos possíveis problemas causados pela ingestão de altas concentrações de sulfato, o valor
limítrofe para águas de abastecimento no Brasil, estabelecido pela Portaria 518/2004 do
Ministério da Saúde, é 250 mg/L. A Resolução 358/2005 do CONAMA (2005) e a
Deliberação Normativa COPAM n° 010/86 (1986) do estado de Minas Gerais limita a
concentração de sulfato para águas doces das Classes 1 e 2 também em 250 mg/L. Esse limite
é definido para o corpo receptor e não para o efluente em si. No Estado de São Paulo, a Lei nº
997 de 31 de maio de 1976, no Artigo 19-A, estabelece a concentração máxima de1,0gSSO42-/L em efluentes líquidos lançados direta ou indiretamente nos corpos receptores,
impondo obrigatoriedade de tratamento às fontes emissoras cujas concentrações de sulfato
ultrapassem este valor determinado. A recomendação dos órgãos fiscalizadores de outros
países é que a concentração de sulfato esteja entre 250 a 500mg/L em águas de abastecimento
e/ou efluentes (INAP, 2003; WHO, 2011).
As tecnologias disponíveis para a remoção do sulfato podem ser classificadas em três
grupos:(i) processos de precipitação química (gesso, etringita e sulfato de bário); (ii)
processos que utilizam membranas ou resina de troca iônica; (iii) redução biológica. As
técnicas que envolvem os processos físico-químicos têm se mostrado ou economicamente
inviáveis (precipitação como etringita e sulfato de bário, uso de membranas e resinas de troca
iônica) ou ineficientes (precipitação como gesso) (SILVEIRAet al., 2008). As aplicações
industriais que utilizam processos com membranas ou resina de troca iônica só se viabilizam
comercialmente pela presença de mercado consumidor para a água tratada, como observado
na África do Sul (HUTTONet al., 2009). Nesse sentido, o tratamento biológico de efluentes
2
industriais contendo metais e sulfato, além de potencialmente atender ao critério de reduzir o
teordo ânionpara concentrações inferiores a 250mg/L, surge como uma alternativa
principalmente por diminuir a produção de lodo e permitir a recuperação de metais de
interesse econômico (como Cu, Ni e Zn).
O interesse pela redução de sulfato em sistemas de tratamento de águas residuárias surgiu a
partir dos problemas causados por esta rota metabólica nos reatores metanogênicos.Muito
conhecimento sobre o processo sulfetogênico foi adquirido entre as décadas de 1970-1980s,
por meio das pesquisas voltadas para a prevenção ou minimização da redução de sulfato
durante o tratamento metanogênico de águas residuárias (ISAet al., 1986; RINZEMA e
LETTINGA, 1988; VISSER et al., 1992; VISSER et al., 1993). Tal interesse compreende
desde a produção de H2S, passando por sua toxicidade, seus efeitos corrosivos, as questões de
odor e de aumento da demanda química de oxigênio (DQO) do efluente, bem como a redução
na qualidade e quantidade de biogás produzido (LENSet al., 2002).
A partir da década de 1990, cresceu o interesse em utilizar a redução de sulfato no tratamento
de efluentes ricos em sulfato, tais como, a drenagem ácida de mina (STUCKIet al., 1993) e
águas da lavagem de gases de combustão (LENSet al., 2002). Por conseguinte, diversos
estudos foram realizados buscando melhorar a eficiência do processo de biorredução de
sulfato em reatores anaeróbios. Além das bactérias redutoras de sulfato (BRS), o consórcio
bacteriano presente no processo de tratamento anaeróbio ainda envolve os grupos de bactérias
acidogênicas, acetogênicas e metanogênicas. Estes grupos bacterianos consomem substrato
orgânico, promovendo uma relação sintrófica ou de competição pelo substrato. Na tabela 1.1
podem ser observadas as reações envolvidas no metabolismo destes grupos de bactérias para
alguns substratos orgânicos.
3
Tabela 1.1 Reações envolvidas no metabolismo de grupos de bactérias durante o processo
anaeróbio.
Equação
∆ G°
(kJ/reação)
Reações de Redução de sulfato
4 + + → + - 151,9
+ → + -47,6
+ , → + + , + , -37,7
+ , → + , + , -27,8
+ , → + + . + , -80,2
Reações Acetogênicas
+ → + + + +76,1
+ → + + +48,3
+ → +
+ + -4,2
Reações Metanogênicas
+ + → + + → + -153,6
-31,0
Reações Homoacetogênicas
+ + → + → !, + , -104,6
-56,5
A concentração (expressa em termos da demanda química de oxigênio – DQO) e a qualidade
do substrato orgânico e a concentração de sulfato no efluente podem ser fatores importantes
para uma eficaz remoção do sulfato. Além do substrato, os principais fatores estudados e que
podem afetar o desempenho da redução biológica do sulfato são: pH, temperatura, razão
DQO/sulfato, tempo de detenção hidráulica (TDH) e o tipo de reator utilizado (SHEORAN et
al., 2010). Em relação ao pH e à temperatura, as condições ideais para o crescimento BRS
são: valores de 5 a 9, e de 20°C a 40°C, respectivamente. Valores de pH e temperatura fora
destas faixas normalmente resultam em redução da atividade sulfetogênica (TANG et al.,
2009).
4
A biorredução do íon sulfato ocorre mediante a oxidação de um composto orgânico
(substrato) pela atividade metabólica das BRS. Portanto, no caso de efluentes industriais com
alta concentração de sulfato, mas que não possuem matéria orgânica em sua constituição,
como no caso do efluente gerado pela DAM, a adição de uma fonte de carbono e elétrons
externa se faz necessária.Por causa da variedade de substratos orgânicos que as BRS podem
utilizar (SHEORANet al., 2010), diversas rotas metabólicas durante a redução do sulfato são
possíveis. Tais rotas são influenciadas pela competição, pelo substrato orgânico, entre as BRS
e outros grupos de micro-organismos envolvidos na digestão anaeróbia. As BRS podem
competir com arqueias metanogênicas por acetato e H2(GUPTAet al., 1994; COLLERAN et
al., 1998); com as bactérias acetogênicas, por propionato ou butirato (O'FLAHERTY et al.,
1998) ou com micro-organismos fermentativos por lactato, glicerol, etanol, sacarose, glicose,
entre outros(SHEORANet al., 2010). A razão DQO/sulfato é um fator determinante do nível
de competição entre estes grupos(CHENet al., 2008) e, portanto, do desempenho dos sistemas
sulfetogênicos. A razão DQO/sulfato foi investigada em diferentes estudos envolvendo
quimiostatos (RENet al., 2007), reatores contínuos (DE SMUL et al., 1999; LOPES et al.,
2010) e em batelada (CAOet al., 2009). Chen e Rim (1991) relataram experimentalmente que
para razões acima de 2,7, as arqueias metanogênicas acetoclásticas predominaram sobres as
BRS acetoclásticas. Por outro lado, em estudos em batelada realizados por Cao et al, (2009),
foi encontrada uma melhor taxa de redução de sulfato para a razão DQO/sulfato igual a 3,0
utilizando lactato como substrato. Considerando apenas a atividade específica das BRS, o
valor ideal para razão DQO/sulfato vai depender do substrato utilizado, da diversidade das
BRS e da espécie predominante na biomassa.Teoricamente, uma razão DQO/sulfato igual a
0,67 contêm doadores de elétrons (DQO) suficientes para remover todo o sulfato, assumindo
que 8 elétrons são transferidos do sulfato para o sulfeto.
5
Avaliando a diversidade de BRS, a melhor razão DQO/sulfato poderá ser definida pela forma
como as BRS oxidam a matéria orgânica. Existem dois grupos diferentes de BRS,
classificados quantoà capacidade de oxidar o substrato. As espécies capazes de oxidar o
substrato orgânico de forma completa até CO2(acetoclásticas) são representadas pelas
espécies Desulfobacter, Desulfobacterium, Desulfonema, Desulfosarcina, Desulfococcus,
Desulfomonile, entre outras.As espécies de BRS que oxidam a matéria orgânica de forma
incompleta até acetato (acetogênicas) são representadas pelos gêneros Desulfobulbus,
Desulfomicrobium, Desulfomonas, Desulfovibrio, entre outros(BARTON, 1995). Todas estas
representantes das BRS são caracterizadas pelo uso de sulfato como aceptor final de elétrons
durante a respiração anaeróbia. Muitos estudos relatam o predomínio de BRS que oxidam
incompletamente o substrato durante a redução do sulfato, resultando em acúmulo do acetato,
devido a ausência da atividade sulfetogênica acetoclástica(KAKSONENet al., 2003; CAO et
al., 2009; CELIS-GARCIA et al., 2009). Portanto, em sistemas com alta concentração de
sulfato, se a matéria orgânica é incompletamente oxidada, menor será o fluxo de elétrons para
a redução do íon sulfato, o que sugere que mais substrato deverá ser adicionado, ou seja, a
razão DQO/sulfato deverá ser maior que a estequiométrica (0.67).
O tipo de reator empregado também é um importante fator no desempenho do processo
sulfetogênico (KAKSONEN e PUHAKKA, 2007). A configuração do reator deve
proporcionar ao sistema capacidade de reter alta quantidade de biomassa, bem com elevada
atividade (sulfetogênica) e alta transferência de massa, ou seja, adequado contato entre
substrato e micro-organismos. Nesse sentido, diferentesconfigurações de biorreator têm sido
usadas para estudar a redução anaeróbia do sulfato como o reator anaeróbio de fluxo
ascendente e manta de lodo (como o UASB), o reator de mistura completa (como o CSTR)
(MOOSAet al., 2002), o de leito empacotado (como o EGSB) (CHANG et al., 2000; JONG e
6
PARRY, 2003), o anaeróbio de batelada sequencial (como o AnSBR)(SARTI e ZAIAT,
2011),
o
compartimentado
(BAYRAKDARet
al.,
2009),
o
de
leito
fluidizado
(FBR)(KAKSONENet al., 2003) e o acoplado a membranas (CHUICHULCHERM et al.,
2001). Nestatese, os reatores contínuos UASB e de Leito Fluidizado (RLF) foram escolhidos
para avaliar a biorredução do sulfato devido às suas diferentes características hidrodinâmicas
e tipos de formação da biomassa.
Os reatores UASB são descritos como de construção simples e de baixo custo de operação,
além de trabalhar com baixos valores de TDH no tratamento de esgotos domésticos. Sendo
que para isto é fundamental a formação de uma biomassa floculenta ou granular (SPEECE,
1983).Contudo, no tratamento de efluentes com altas cargas de sulfato, o que se observa é a
ruptura (ou a não formação) destes grânulos (SCHMIDT e AHRING, 1996).
Consequentemente, para evitar que células de bactérias sejam lavadas do reator, é necessário
alto tempo de detenção hidráulica (TDH)(LENSet al., 2002).Além disso, as condições de
mistura no reator devem proporcionar adequado contato entre a biomassa e o substrato. Como
nos reatores sulfetogênicos, a degradação da matéria orgânica é realizada, sobretudo pelas
BRS, convertendo sulfato a sulfeto dissolvido, nestes biorreatores,há pouca formação de
biogás, o que está diretamente relacionadoà baixa atividade metanogênica (inibição pela
toxicidade do íon sulfeto). Diante do fato de que a evolução do biogás tem sido associada à
diminuição da resistência à difusão no lodo granular (LENSet al., 2002), os reatores
sulfetogênicos com biomassa em suspensão operam a menores taxas de transferência de
massa, o que resulta em menores eficiências de redução de sulfato. Nesses reatores, asformas
mais usuaispara aumentar a taxa de transferência de massa são; (i) elevar a velocidade
superficial do líquido no reator(OMILet al., 1996), (ii) injeção de gás inerte no leito do lodo
granular (LENS et al., 2003). Além disso, uma hipótese que se levanta nesta tese é de que a
7
recirculação da biomassa pode ser um recurso eficiente em reatores sulfetogênicos por
aumentar a velocidade superficial do líquido e, portanto, melhorar a condição de mistura no
reator UASB.
A aplicação do reator UASB em processos sulfetogênicos é estudada em função de diferentes
parâmetros como, o tipo de substrato, a competição entre BRS e outros micro-organismos, a
razão DQO/sulfato e velocidade ascendente do fluxo. A Tabela 1.2 apresenta as condições
operacionais e o desempenho de reatores UASB,em estudos de redução de sulfato, realizados
nos últimos 10 anos.
8
Tabela 1.2.Desempenho de vários reatores UASB usados para a redução de sulfato.
Sulfato
Referência
Lopes et al.(2010)
Características do sistema
Substrato
T(ºC)
COA
(KgDQO/m3.d)
Razão
DQO/SO42-
55
4.0
4
5.0
38
5.25-0.315
7-1.2
pH
Afluente
(kgSO42-/m3.d)
Redução
(%)
Residual
(mg.L-1)
0.98
50-70%
120-200
5.0
0.273-4.55
84-98%
50-430
TRH -10h; com recirculação
Sacarose
TRH -0.6-10d; com
Efluente da
recirculação
suinocultura
Poinapen et al (2009)
Lodo primário*
35
3.45
1.44
6.0
2.4
92%
144
Poinapen et al. (2009)
Lodo primário*
20
1.975
1.75
6.0
1.7
93%
109
Sacarose
55
4.76
1
6.0
4.76
95%
100
Etanol
25
-
0.75
7.0
0.53
38%
248
30
4.26
4.85
7.0
0.88
90-95%
68-151
Melaço
35
0.65
2
7.0
0.32
>80%
99
Kosinska and Miskiewicz
(2009)
Lopes et al., (2007)
Gonçalves et al. (2007)
e injeção de N2
TRH -18h; semrecirculação
Mohan et al. (2005)
TRH -37h; sem recirculação
Efluente
industrial**
Shayegan et al. (2005)
TRH -37h; velocidades-0.5 e
1m/h; sem recirculação
Kaksonen et al. (2003)
Lactato
35
-
-
3.0
2.46
75%
402
Vallero et al. (2003)
TRH -7.5h; com recirculação
Metanol
55
18.56
5
7.0
3.84
99%
12
*Lodo não estabilizado do tratamento de água residuária de um sistema de Lodo Ativado.
**Mistura de substâncias biodegradáveis da indústria química.
9
Além dos fatores acima citados (natureza do substrato e razão DQO/sulfato) que influenciam
a redução do sulfato, outro aspecto importante é a taxa de crescimento específica
relativamente baixa das BRS (para Desulfovibrio sp, são citados valores de µ maxem torno de
0,25h-1)(ZELLNER et al., 1994). O problema do crescimento de culturas microbianas que
apresentam baixas taxas de crescimento, é resolvido pelo uso de reatores que utilizam
biomassa imobilizada como os reatores de leito fixo, por exemplo(HAMMACK e
EDENBORN, 1992; GROUDEVA e GROUDEV, 1997; HAMMACK e DIJKMAN, 1999).
Entretanto, estes reatores estão sujeitos a entupimentos e problemas de transferência de
massa, o que diminui seus níveis de eficiência a longo prazo(SOMLEV e BANOV, 1998).
Uma alternativa é o uso de reatores de leito fluidizado (KAKSONENet al., 2006)que possuem
eficiente retenção de biomassa, permitindo rápida transferência de massa e, portanto, altas
taxas de redução (KAKSONEN e PUHAKKA, 2007). Nos reatores de leito fluidizado, a
biomassa é retida em um material suporte inerte, que é fluidizado por reciclo do efluente ou
por fluxo de gás. O carvão ativado foi escolhido nestatese por fornecer alta área superficial
específica para a formação do biofilme (SUTTON e MISHRA, 2004). A tabela1.3reúne
alguns estudos onde o desempenho de reatores de leito fluidizado, com diferentes materiais
suportes foi estudado. Pode-se observar que altas taxas volumétricas de redução de sulfato
têm sido reportadas nestes reatores.
10
Tabela 1.3.Desempenho de vários reatores de leito fluidizado usados para a redução do sulfato.
Referências
Características do sistema
Substrato
T
(ºC)
Razão
DQO/SO42-
Sulfato
pH
Afluente
(kgSO42-/m3.d)
Redução (%)
Residual
(mg.L-1)
Nevatalo et al. (2010)
TRH-8h
Fluidização-20%
Etanol + metais
35
0.67-0.75
4
4.55
47
800
Celis-García et al.. (2009)
TRH - 2d
Fluidização -25%
Polietileno
Etanol
25
0.6
6.5
1.66
28
2400
Sahinkaya et al. (2007)
TRH -1d
Mineral silicatado
Etanol
8
0.74
6.5
1.63
23
1255
TRH 0.67 – 1d
Fluidização 25%Polietileno
Lactato/Propionato/
Butirato
30
0.67 – 1.25
7
2.0 – 7.3
73-79
175-1024
Sahinkaya et al.(2011)
TRH -1d
Mineral silicatado
Fluidização-25%
Etanol
65
0.67
-
1.5
70
450
TRH -16h
Fluidização -20%
Material silicatado
Lactato e Etanol
35
0.67
5-2.5
1.49 - 3.3
77-95(lactate)
60-75(Ethanol)
-
TRH -6.5h
Fluidização -20%
Mineral silicatado
Etanol
35
0.67
3
7.7
57
894
TRH -5h e 55h
Etanol
30
2
7
11.9
95(55h)
65 (5h)
161(55h)
1171 (5h)
Celis-García et al. (2007)
Nagpal et al. (2000)
11
A principal barreira para a aplicação da tecnologia de biorredução de sulfato para o
tratamento de efluentes com alta carga de sulfato é o custo da fonte de carbono e elétrons
(INAP, 2003). Tanto a seleção de uma fonte de carbono adequada quanto a quantidade do
substrato utilizada estão relacionadas à sustentabilidade dos reatores sulfetogênicos. Dos mais
de 34 substratos identificados como utilizáveis pelas BRS (SHEORANet al., 2010), os ácidos
graxos voláteis (acetato, propionato e butirato) e os ácidos graxos produtos intermediários da
fermentação (lactato, piruvato e malato) estão entre as principais fontes de carbono e
elétrons(LIAMLEAM e ANNACHHATRE, 2007). Em termos de energia disponível e
biomassa produzida, o lactato é um doador de elétrons superior ao etanol, ao acetato, ao
propionato e ao ácido acético (NAGPAL, CHUICHULCHERM, LIVINGSTONet al., 2000),
além de poder ser utilizado por uma grande variedade de espécies de BRS (BARTON, 1995).
Portanto, é uma escolha lógica como controle positivo (SHEORANet al., 2010), nos estudos
de seleção de diferentes fontes de carbono para crescimento de BRS. Porém, seu uso em
escala real é economicamente inviável. Diversas fontes de carbono mais baratas têm sido
propostas tais como, feno, alfafa, lascas de madeira, estrume de animais, melaço e lodo de
esgoto (LIAMLEAM e ANNACHHATRE, 2007). Considerando que as BRS não são capazes
de utilizar diretamente substratos orgânicos complexos, a presença de micro-organismos
capazes de degradar estes compostos a moléculas mais simples é essencial para o desempenho
da redução de sulfato (TANGet al., 2009).
Com o desenvolvimento da indústria de biocombustíveis, particularmente de biodiesel, a
disponibilidade de glicerol bruto será cada vez maior e, portanto, novas formas para o uso
deste resíduo industrial são necessárias. A alta energia contida no glicerol faz com que seja
um interessante substrato para a digestão anaeróbia (KOLESÁROVÁet al., 2011). Estudos
recentes o têm utilizado nestarota como fonte única de carbono ou combinado com diferentes
12
substratos para a produção de biogás em reatores metanogênicos (YANGet al., 2008;
FOUNTOULAKIS e CZACZYK, 2009; LOPÉZ et al., 2009; ALVAREZ et al., 2010).
Entretanto, na redução de sulfato, poucos estudos que utilizam o glicerol como fonte de
carbono e elétrons têm sido relatados(QATIBI, 1990; DINKELet al., 2010). As BRS relatadas
capazes de oxidar glicerol são todas do gênero Desulfovibrio(KREMER e HANSEN, 1987;
QATIBIet al., 1991) o que implica que o glicerol é degradado incompletamente a acetato.
Entretanto, o glicerol pode ser fácil e rapidamente fermentado por outras bactérias, o que
reduz o pH dos reatores e inibe a atividade sulfetogênica. Nesse sentido, Qatibi et al. (1991)
descreveram em estudo com cultura mista de micro-organismos, em que as BRS competem
com bactérias fermentativas pelo glicerol e sugeriram que a redução de sulfato ocorre
acoplada à oxidação de produtos da fermentação do glicerol. Assim, a presença de microorganismos fermentativos parece ser essencial para a redução do sulfato a partir da
degradação do glicerol. Nesta tese, postula-se a hipótese de que o glicerol pode ser substrato
para as BRS promovendo a redução do sulfato como uma fonte viável de matéria orgânica.
Esse conjunto de aspectos da redução biológica de sulfato, sucintamente descritos, é o suporte
para o desenvolvimento dessa tese. Seus objetivos e a organização geral do texto são
apresentados a seguir.
1.2. Objetivos e organização da tese
A configuração do reator, as condições operacionais aplicadas e o substrato orgânico são
importantes ferramentas para avaliar a aplicabilidade da tecnologia anaeróbia para a redução
biológica do sulfato. Neste estudo, estas ferramentas foram comparadas em dois reatores
anaeróbios, UASB e de Leito Fluidizado, com os seguintes objetivos específicos:
13
•
Definir as rotas metabólicas predominantes no reator UASB, em função dadiversidade
de micro-organismos presentes na biomassa (Capítulo III).
•
Identificar os principais grupos de micro-organismos presentes na biomassa dos
reatores através de técnicas moleculares (Capítulo II, III e IV).
•
Avaliar o efeito do aumento da carga orgânica na competição entre as bactérias
redutoras de sulfato e fermentativas no reator UASB (Capítulo III).
•
Avaliar o efeito da recirculação da biomassa no reator UASB na redução do sulfato
(Capítulo III).
•
Descrever e comparar as rotas metabólicas predominantes em um reator de leito
fluizidado (RLF)sob alta carga de sulfato, utilizando lactato e/ou glicerol puro como
fonte de carbono e elétrons (Capítulo IV).
•
Comparar as eficiências de redução do sulfato utilizando um substrato controle
(lactato) e um alternativo (glicerol) no reator RLF (Capítulo IV).
•
Determinar os parâmetros para a redução de sulfato nos reatores contínuos, UASB e
RLF (Capítulo II).
•
Comparar as taxas específicas de redução de sulfato entre o reator com imobilização
de biomassa (Leito Fluidizado) e o reator com biomassa dispersa (UASB) (Capítulo
II).
Neste sentido, esta tese de doutorado foi organizada em capítulos, sendo que no CAPÍTULO
2 estão resumidos os principais resultados obtidos durante o monitoramento de um reator com
biomassa dispersa, reator UASB, e um com biomassa aderida, reator de leito fuidizado (RLF),
no trabalho intitulado “PERFORMANCE OF CONTINUOUS BIOREACTORS FOR
SULFATE REDUCTION AIMING AT UTILIZING GLYCEROL AS CARBON
SOURCE”.O trabalho comparou o desempenho dos reatores em função de suas
14
configurações e condições operacionais e discutiu a eficiência de redução do sulfato,
selecionando o melhor reator para avaliar a eficiência da biorredução do sulfato utilizando o
glicerol como fonte de carbono e elétrons. Este artigo foi submetido ao periódico
“Biochemical Engineering Journal”.
No CAPÍTULO 3, serão discutidos os resultados obtidos durante o monitoramento do reator
UASB, no trabalho intitulado:“IMPLICATIONS OF VOLATILE FATTY ACID
PROFILE ON THE METABOLIC PATHWAY DURING A CONTINUOUS
SULPHATE REDUCTION”. Neste artigo,a eficiência da biorredução do sulfato no reator
UASB
foi detalhadamente discutidasob diferentes cargas orgânicas e
condições
hidrodinâmicas (sem e com recirculação). As principais rotas metabólicas foram propostas
para o processo de redução do sulfato durante a degradação do lactato em função da
diversidade microbiana presente no reator. Este artigo foi publicado na revista “Jornal
Environmental Management”.
No CAPÍTULO 4, será apresentado o trabalho intitulado “GLYCEROL AS A SINGLE
ELECTRON DONOR FOR SULFATE REDUCTION ON FLUIDIZED BED
REACTORS”. Neste trabalho, altas cargas de sulfato foram aplicadas em um reator de leito
fluidizado com carvão ativado como material suporte. Primeiramente, a consolidação de uma
biomassa predominantemente sulfetogênica foi alcançada, tendo o lactato como substrato.
Além disso, um substrato alternativo, o glicerol, foi avaliado como fonte de carbono e
elétrons, em função das espécies de BRS presentes no reator. As rotas metabólicas de
degradação do lactato e glicerol foram propostas e comparadas. Este artigo foi submetido à
revista “Journal of Hazardous Materials”.
15
1.3. Referências
AKCIL, A. e S. KOLDAS. Acid Mine Drainage (AMD): causes, treatment and case studies.
Journal of Cleaner Production, v.14, n.12, p.1139-1145. 2006.
ALVAREZ, J. A., et al. A methodology for optimising feed composition for anaerobic codigestion of agro-industrial waste. Bioresource Technology, v.101, n.4, p.1153-1158. 2010.
BARTON, L. L. Sulfate-reducing bacteria. New York: Plenum Press, v.8. 1995. 336 p.
(Biotechonology Handbooks)
BAYRAKDAR, A., et al. Performance of sulfidogenic anaerobic baffled reactor (ABR)
treating acidic and zinc-containing wastewater. Bioresource Technology, v.100, n.19, p.43544360. 2009.
CAO, J., et al. Precipitation of valuable metals from bioleaching solution by biogenic sulfides.
Minerals Engineering, v.22, n.3, p.289-295. 2009.
CELIS-GARCIA, L., et al. Characterization of sulfate-reducing bacteria dominated surface
communities during start-up of a down-flow fluidized bed reactor. Journal of Industrial
Microbiology & Biotechnology, v.36, n.1, p.111-121. 2009.
CELIS-GARCÍA, L. B., et al. Performance of a down-flow fluidized bed reactor under sulfate
reduction conditions using volatile fatty acids as electron donors. Biotechnology and
Bioengineering, v.97, n.4, p.771-779. 2007.
CHANG, I. S., et al. Biological treatment of acid mine drainage under sulphate-reducing
conditions with solid waste materials as substrate. Water Research, v.34, n.4, p.1269-1277.
2000.
CHEN, W. H. e J. M. RIM.Competition and inhibition of sulfate reducers and methane
producers in anaerobic treatment.Water Science Technology, v.23, p.1259-1264. 1991.
CHEN, Y., et al. Inhibition of anaerobic digestion process: A review. Bioresource
Technology, v.99, n.10, p.4044-4064. 2008.
CHUICHULCHERM, S., et al. Treatment of metal-containing wastewaters with a novel
extractive membrane reactor using sulfate-reducing bacteria.Journal of Chemical Technology
& Biotechnology, v.76, n.1, p.61-68. 2001.
COLLERAN, E., et al. Full-scale and laboratory-scale anaerobic treatment of citric acid
production wastewater.Biodegradation, v.9, n.3, p.233-245. 1998.
CONAMA. Conselho Nacional do Meio Ambiente - Resolução
<http://www.mma.gov.br/port/conama/res/res05/res35705.pdf> 2005.
n°
357/2005:
COPAM. Deliberação Normativa N° 10, de 10 de dezembro de 1986. C. P. Ambiental: Diário
do Executivo-"Minas Gerais" 1986.
16
DE SMUL, A., et al. Effect of COD to sulphate ratio and temperature in expanded-granularsludge-blanket reactors for sulphate reduction. Process Biochemistry, v.34, n.4, p.407-416.
1999.
DINKEL, V., et al. Kinetics of anaerobic biodegradation of glycerol by sulfate-reducing
bacteria.Applied Biochemistry and Microbiology, v.46, n.7, p.712-718. 2010.
FOUNTOULAKIS, M. S. e K. CZACZYK. Enhanced methane and hydrogen production
from municipal solid waste and agro-industrial by-products co-digested with crude glycerol.
Bioresource Technology, v.100, p.3043-3047. 2009.
GONÇALVES, M. M. M., et al. Heavy metal removal from synthetic wastewaters in an
anaerobic bioreactor using stillage from ethanol distilleries as a carbon source. Chemosphere,
v.69, n.11, p.1815-1820. 2007.
GROUDEVA, V. I. e S. N. GROUDEV. Bioremediation of Acid Drainage Waters from a
Cooper Mine. In: International Biohydrometallurgy Symposium IBS-97. Sidney, Australia,
1997. p. 50.
GUPTA, A., et al. Methanogenesis and sulfate reduction in chemostats.Kinetic studies and
experiments.Water Research, v.28, n.4, p.781-793. 1994.
HAMMACK, R. W. e H. DIJKMAN. The Application of Bacterial Sulfate Reduction
Treatment to Severely Contaminated Mine Waters: Results of Three Years of Pilot Plant
Testing. In: Copper 99 - Cobre 99. Phoneix, Arizona, USA: TMS, 1999. p. 97-111.
HAMMACK, R. W. e H. M. EDENBORN.The Removal of Nickel from Mine Waters Using
Bacterial Sulfate Reduction.Applied Microbiology and Biotechnology, v.37, n.5, Aug, p.674678. 1992.
HUTTON, B., et al. Operating and maintenance experience at the eMalahleni Water
Reclamation Plant. In: International Mine Water Conference. Pretoria, South Africa, 2009. p.
415-430.
INAP.Treatment of sulphate in mine effluentes.International network for acid prevention,
p.129. 2003
ISA, Z., et al. Sulfate Reduction Relative to Methane Production in High-Rate Anaerobic
Digestion: Technical Aspects. Appl. Environ. Microbiol., v.51, n.3, March 1, 1986, p.572579. 1986.
JONG, T. e D. L. PARRY. Removal of sulfate and heavy metals by sulfate reducing bacteria
in short-term bench scale upflow anaerobic packed bed reactor runs. Water Research, v.37,
n.14, p.3379-3389. 2003.
KAKSONEN, A., et al. Performance and Ethanol Oxidation Kinetics of a Sulfate-Reducing
Fluidized-Bed Reactor Treating Acidic Metal-Containing Wastewater.Biodegradation, v.14,
n.3, p.207-217. 2003.
17
KAKSONEN, A. H., et al. The performance, kinetics and microbiology of sulfidogenic
fluidized-bed treatment of acidic metal- and sulfate-containing wastewater.Hydrometallurgy,
v.83, n.1-4, p.204-213. 2006.
KAKSONEN, A. H. e J. A. PUHAKKA. Sulfate Reduction Based Bioprocesses for the
Treatment of Acid Mine Drainage and the Recovery of Metals. Engineering in Life Sciences,
v.7, n.6, p.541-564. 2007.
KOLESÁROVÁ, N., et al. Utilization of Biodiesel By-Products for Biogas Production.
Journal of Biomedicine and Biotechnology, v.2011, p.1-15. 2011.
KOSINSKA, K. e T. MISKIEWICZ.Performance of an anaerobic bioreactor with biomass
recycling, continuously removing COD and sulphate from industrial wastes.Bioresource
Technology, v.100, n.1, p.86-90. 2009.
KREMER, D. R. e T. A. HANSEN.Glycerol and dihydroxyacetone dissimilation in
Desulfovibrio strains.Archives Microbiology, v.147, p.249-256. 1987.
LENS, P., et al. Perspectives of sulfate reducing bioreactors in environmental
biotechnology.Reviews in Environmental Science and Biotechnology, v.1, n.4, p.311-325.
2002.
LENS, P. N., et al. Sulfate reducing and methane producing bacteria in aerobic wastewater
treatment systems.Water Research, v.29, n.3, p.871-880. 1995.
LENS, P. N. L., et al. Use of Sulfate Reducing Cell Suspension Bioreactors for the Treatment
of SO2 Rich Flue Gases.Biodegradation, v.14, n.3, p.229-240. 2003.
LENS, P. N. L., et al. Biotechnological treatment of sulfate-rich wastewater.Critical reviews
in Environmental Science and Technology, v.28, n.1, p.41-88. 1998.
LIAMLEAM, W. e A. P. ANNACHHATRE.Electron donors for biological sulfate
reduction.Biotechnology Advances, v.25, n.5, p.452-463. 2007.
LOPES, S. I. C., et al. Sulfate reduction during the acidification of sucrose at pH 5 under
thermophilic (55 °C) conditions. II: Effect of sulfide and COD/ SO 4 2 - ratio. Bioresource
Technology, v.101, n.12, p.4278-4284. 2010.
LOPES, S. I. C., et al. Effect of COD/SO42- ratio and sulfide on thermophilic (55°C) sulfate
reduction during the acidification of sucrose at pH 6.Water Research, v.41, n.11, p.23792392. 2007.
LOPÉZ, J. A. S., et al. Anaerobic digestion of glycerol derived from biodiesel manufacturing.
Bioresource Technology, v.100, p.5609-5615. 2009.
MOHAN, S. V., PRASAD, K.K., RAO, N.C., BHASKAR, Y.V., BABU, V.L.,
RAJAGOPAL, D. AND SARMA, P.N. Biological treatment of low-biodegradable composite
chemical wastewater using upflow anaerobic sludge blanket (UASB) reactor. Journal of
scientific and industrial research, v.64, p.771-777. 2005.
18
MOOSA, S., et al. A kinetic study on anaerobic reduction of sulphate, Part I: Effect of
sulphate concentration. Chemical Engineering Science, v.57, n.14, p.2773-2780. 2002.
NAGPAL, S., et al. Ethanol utilization by sulfate-reducing bacteria: An experimental and
modeling study. Biotechnology and Bioengineering, v.70, n.5, p.533-543. 2000.
NAGPAL, S., et al. Microbial sulfate reduction in a liquid–solid fluidized bed reactor.
Biotechnology and Bioengineering, v.70, n.4, p.370-380. 2000.
NEVATALO, L. M., et al. Biological hydrogen sulfide production in an ethanol lactate fed
fluidized-bed bioreactor. Bioresource Technology, v.101, n.1, p.276-284. 2010.
O'FLAHERTY, V., et al. Long-term competition between sulphate-reducing and methaneproducing bacteria during full-scale anaerobic treatment of citric acid production
wastewater.Water Research, v.32, n.3, p.815-825. 1998.
OMIL, F., et al. Effect of upward velocity and sulphide concentration on volatile fatty acid
degradation in a sulphidogenic granular sludge.Process Biochemistry 31 (1996) 699-710.
1996.
POINAPEN, J., et al. Biological sulphate reduction with primary sewage sludge in an upflow
anaerobic sludge bed (UASB) reactor - Part 2: Modification of simple wet chemistry
analytical procedures to achieve COD and S mass balances. Water SA, v.35, p.535-542. 2009.
POINAPEN, J., et al. Biological sulphate reduction with primary sewage sludge in an upflow
anaerobic sludge bed (UASB) reactor - Part 3: Performance at 20ºC and 35ºC. Water SA,
v.35, p.543-552. 2009.
QATIBI, A. I. Fermentation du lactate, du glycerol et des diols par les bactéris sulfatoréductrices du genere Desulfovibrio. Microbiology, University of Provence, France, 1990.165
p.
QATIBI, A. I., et al. Desulfovibrio alcoholovorans sp. nov., a sulfate-reducing bacterium able
to grow on glycerol, 1,2-and 1,3-propanediol. Microbiology, v.155, p.143-148. 1991.
REN, N.-Q., et al. Effects of COD/SO42- Ratios on an Acidogenic Sulfate-Reducing Reactor.
Industrial & Engineering Chemistry Research, v.46, n.6, 2007/03/01, p.1661-1666. 2007.
RINZEMA, A. e G. LETTINGA.Anaerobic treatment of sulfate containing wastewater. In:
W. D.L. (Ed.). Biotreatment systems. Bonda Raton, FL: CRC Press, 1988. p.65 - 109
SAHINKAYA, E., et al. Sulfidogenic fluidized-bed treatment of metal-containing wastewater
at 8 and temperatures is limited by acetate oxidation. Water Research, v.41, n.12, p.27062714. 2007.
SAHINKAYA, E., et al. Sulfidogenic fluidized bed treatment of real acid mine drainage
water. Bioresource Technology, v.102, n.2, p.683-689. 2011.
19
SARTI, A. e M. ZAIAT.Anaerobic treatment of sulfate-rich wastewater in an anaerobic
sequential batch reactor (AnSBR) using butanol as the carbon source.Journal of
Environmental Management, v.92, n.6, p.1537-1541. 2011.
SCHMIDT, J. E. e B. K. AHRING.Granular sludge formation in upflow anaerobic sludge
blanket (UASB) reactors.Biotechnology and Bioengineering, v.49, n.3, p.229-246. 1996.
SHAYEGAN, J., et al. The effect of influent COD and upward flow velocity on the behaviour
of sulphate-reducing bacteria.Process Biochemistry, v.40, n.7, p.2305-2310. 2005.
SHEORAN, A. S., et al. Bioremediation of acid-rock drainage by sulphate-reducing
prokaryotes: A review. Minerals Engineering, v.23, n.14, p.1073-1100. 2010.
SILVEIRA, A. N., et al. Técnicas para tratamento e alternativas de reúso de águas ácidas de
minas de carvão. In: VI Simpósio Internacinal de Qualidade Ambiental da Associação
Brasileira de Engenharia Sanitária (ABES). Porto Alegre. 26 - 28, maio, 2008, 2008. p. 1-8.
SOMLEV, V. e M. BANOV. Three stage process for complex biotechnological treatment of
industrial wastewater from uranium mining. Biotechnology Techniques, v.12, n.8, Aug, p.637639. 1998.
SPEECE, R. E. Anaerobic biotechnology for industrial wastewaster treatment.Environmental
Science Technology v.17, n.9, p.416 - 427. 1983.
STUCKI, G., et al. Biological sulfuric acid transformation: Reactor design and process
optimization. Biotechnology and Bioengineering, v.41, n.3, p.303-315. 1993.
SUTTON, P. M. e P. N. MISHRA. Activated carbon based biological fluidized beds for
contaminated water and wastewater treatment: a state-of-the-art review. Water Science and
Technology, v.29, p.309-317. 2004.
TANG, K., et al. Bacteria of the sulphur cycle: An overview of microbiology, biokinetics and
their role in petroleum and mining industries. Biochemical Engineering Journal, v.44, n.1,
p.73-94. 2009.
VALLERO, M. V. G., et al. Effect of sulfate on methanol degradation in thermophilic (55 °C)
methanogenic UASB reactors. Enzyme and Microbial Technology, v.32, n.6, p.676-687.
2003.
VISSER, A., et al. Granulation and immobilisation of methanogenic and sulfate-reducing
bacteria in high-rate anaerobic reactors.Journal Applied Microbiology and Biotechnology,
v.40, p.575-581. 1993.
VISSER, A., et al.The anaerobic treatment of synthetic sulfate containing wastewater under
thermophilic (55°C) conditions.Water Science Technology, v.25, p.193-202. 1992.
WHO.Guidelines for drinking-water quality.Genebra. 1: 564 p. 2011.
20
YANG, Y., et al. Biodegradation and methane production from glycerol-contaning synthetic
waste with fixed-bed bioreactor under mesophilic and thermophilic anaerobic conditions
Process Biochemistry, v.43, p.362-367. 2008.
ZELLNER, G., et al. Degradation of Lactate by an Anaerobic Mixed Culture in a FluidizedBed Reactor. Water Research, v.28, n.6, p.1337-1340. 1994.
21
CAPÍTULO 2
PERFORMANCE OF CONTINUOUS BIOREACTORS FOR SULFATE REDUCTION
AIMING AT UTILIZING GLYCEROL AS CARBON SOURCE
Abstract
Reactor hydrodynamics is important for sulfidogenesis because sulfate reduction bacteria
(SRB) do not granulate easily. In this work, the sulfate reduction performance of an Upflow
Anaerobic Sludge Blank (UASB) reactor was compared to that of a fluidized bed reactor
(FBR). Organic loading, sulfate reduction, COD removal and mixing conditions were the
main parameters monitored during lactate and glycerol degradation. The UASB reactor
showed a maximum volumetric sulfate reduction specific rate of 1.3g/(L-1.d -1 (66% removal)
working without recirculation, and this was a result of competition between fermentative
bacteria and SRB for the available substrate. Biomass recirculation considerably enhanced the
sulfate reduction specific rate to 0.089±0.014g/(gSSV.d) (89% reduction), for a COD/sulfate
mass ratio value of 2.5±0.2, whereas values twice as higher were achieved in the FBR treating
either lactate (0.191±0.016g/(gSSV.d)) or glycerol (0.172±0.010 g/(gSSV.d)). Sulfate
reduction in the presence of pure glycerol as an organic substrate produced a smaller residual
COD (1700mg/L) than that produced with lactate (2500mg/L) at the same COD/sulfate mass
ratio of 2.5. It was estimated that 50% of glycerol degradation was due to sulfate reduction
and 50% to fermentation, which was supported by the presence of butyrate in the FBR
effluent. The UASB reactor was unable to produce an effluent with sulfate concentrations
below 250mg/L due to poor mixing conditions. Conversely, the FBR consistently ensured
residual sulfate concentrations below this target value.
Keywords: sulfate reduction; upflow anaerobic sludge blanket (UASB) reactor; fluidized bed
reactor; glycerol; reactor selection.
22
2.1. Introduction
Treatment of sulfate-containing effluents is a major issue for both mining, metallurgical and
chemical industries, due to a frequently large anion content, which can reach 8g/L [1]. The
reasons for such contamination are the widespread use of sulfuric acid in chemical and
metallurgical industries, in addition to the natural oxidation of sulfide minerals in mining
operations.
Sulfate is not a very toxic compound, but above 600mg/L in drinking water, it usually has
laxative effects. Therefore, the World Health Organization (WHO) does not establish a
guideline value for sulfate and only recommends that authorities should be notified when the
anion concentration is above 500mg/L in drinking water. On the other hand, since the
presence of sulfate in concentrations higher than 250mg/L may affect acceptability of
drinking water, this concentration is usually taken as target from a water quality perspective.
Regarding wastewaters, most countries do not specify a value for sulfate, but maximum TDS
(total dissolved solids) limits are usually set implying that sulfate concentrations must comply
with such limits [1]. Overall, discharge limits varying between 250mg/L and 500mg/L are
common place in mining countries, requiring effluent treatment if sulfate concentrations are
above such threshold value [1, 2].
Among the high-rate anaerobic reactors applied to sulfate reduction, the upflow anaerobic
sludge blanket (UASB) reactor [3-9] and the fluidized bed reactor (FBR) [4, 10-13] are the
most studied. Ideally, both reactors must ensure a high concentration of active biomass, along
with good mixing conditions so that high performances can be achieved [14-16]. Furthermore,
in the case of UASB reactors, the residence time must be larger than the generation time to
23
avoid microorganism washout [17, 18]. Overall, the performance of anaerobic reactors
treating high sulfate loading rates (SLR) is defined by: (i) substrate type [19]; (ii) COD/sulfate
ratio [8, 20]; (iii) inoculum source and enrichment procedure [21]; (iv) pH values [22]; (v)
competition among it different groups of microorganisms [23, 24] and reactor configuration
[13, 25]. Moreover, competition between sulfate-reducing bacteria (SRB) and methaneproducing microorganisms (MPM) in anaerobic reactors is well documented [26-28], but the
fermentative metabolism, although having an important contribution to organic matter
oxidation [24, 29, 30], is less discussed in the context of continuous sulfate reduction.
It is also worth considering that the main barriers for the widespread implementation of a
biological alternative for sulfate removal are both the cost of organic matter and the need of
downstream COD removal. An alternative organic substrate could be crude glycerol (gphase). This is a by-product of biodiesel production that contains approximately 50-60%
glycerol, 12-16% alkali soaps and hydroxides, 15-18% methyl-ethers, 8-12% methanol and 23% water. With the development of the biodiesel industry, a surplus of crude glycerol is
foreseen, but it has been tested mostly as a substrate for methane production [31-34] and only
a few studies addressed glycerol application as a potentially inexpensive carbon and electron
source for SRB growth [29, 35]. Therefore, this work initially sought to comprehensively
compare the performance of two different bioreactors treating sulfate-laden waters: (i) an
UASB reactor, which has a simple and inexpensive design and does not require a supporting
material for bacterial growth; (ii) a fluidized bed reactor in which activated carbon was
utilized as support. The second goal was to investigate the use of pure glycerol as carbon
source for sulfate reduction in the fluidized bed reactor as a preliminary step before
investigating the use of crude glycerol.
24
2.2. Materials and methods
2.2.1. Bioreactors
Two lab-scale bioreactors were projected and assembled as shown in figure 2.1. Both reactors
were placed inside a fume hood in a temperature-controlled room, whereby the temperature
was maintained at 25±2ºC. Peristaltic pumps fed Postgate C medium supplemented with
sulfate into both reactors. The UASB reactor had a total volume of 3.0L and contained three
sampling ports (a, b and c) and a gas outlet at its top. Peristaltic pump was used for biomass
recirculation by port c, as detailed in Table 2.1.
Figure 2.1 Pictures of the two lab-scale reactors, UASB and FBR. Port c in the UASB reactor
was utilized for biomass recirculation during phase VII. In the FBR biomass was performed
from point g.
25
Table 2.1.Characteristics and operating conditions of the reactors studied.
Parameters
Volume (L)
Flow rate (L/h)
Hydraulic retention time (h)
UASB
3.0
0.125-0.167±0.01
18 (phase VI)
FBR
1.3
0.13±0.01
10±1 (all phases)
24±1 (phases I-V; VII)
Recirculation rate (L/h)
Temperature
Carrier material
Fluidization (%)
12
18 ±1 (phase VI)
25±2
-
166
25±2
activated carbon
86
The total volume of the FBR was 1.3 liter. Three sampling ports (D, E and F), a gas outlet
(G), a feed tank, as well as an effluent tank completed the system. Activated carbon was used
as the biomass carrier material (150g; 2.1mm mean diameter; density: 1.63g/cm3) and it was
fluidized by means of flow recirculation by a second pump with the flow rate set at 166L/h
(table 2.1). This resulted in a upflow velocity of 75.0m/h and 86% of bed expansion (table
2.1). For fluidization, the effluent from the outlet port (g) was recycled in to the system.
2.2.2. Microorganisms and reactor start-up
The original inoculum (granular sludge) was obtained from an UASB reactor (real scale)
treating domestic wastewater. Enrichment of sulfate-reducing bacteria was performed in a
batch reactor (5 liters) for about 500 days with Postgate C mineral medium containing 0.5g/L
KH2PO4; 1.0g/L NH4Cl; 0.06g/L MgSO4.7H2O; 0.1g/L FeSO4.7H2O; 0.25g/L yeast extract;
2.96g/L Na2SO4; and 3.76g/L lactate as carbon and electron source.
The time diagram depicted in figure 2.2, shows the experimental conditions applied in each
reactor. Postgate C medium, with variable sulfate and lactate concentrations, was applied for
26
growth. During the FBR operation, sulfate concentration was kept at 2.0 gSO42-/L in phases I
and II, while in phases III and IV, the COD was set at 5.0g/L. The optimum COD/sulfate ratio
(2.5) was applied during phase V, aiming at preparing the FBR for a substrate change (from
lactate to glycerol). Phase VI was run with glycerol as the only carbon source, since it
replaced lactate in the Postgate C medium. Similarly, the operational conditions for the UASB
reactor were as follow: reactor start-up during phases I and II; COD increasing from
3.6gCOD/L to 6.0gCOD/L in phases III to V; flow rate change from 0.125 L/h to 0.167L/h
(HRT reduced from 24h to 18h) (phase VI) and effluent recirculation during phase VII.
Figure 2.2.Time diagram showing experimental conditions applied in both the UASB reactor
and the FBR. Inside each box is depicted the COD/sulfate mass ratio. When there was a
change on the COD or sulfate loading the other parameter was kept constant. During phase VI
(UASB reactor), the change on the COD/Sulfate ratio was due to different flow rate applied.
The effects of COD/sulfate mass ratio, upflow velocities (UASB) and substrate type (FBR) in
the performance of both reactors were assessed. The accomplish this the reactor effluents
were analyzed twice a week for total and filtered chemical oxygen demand (COD), sulfate,
alkalinity, volatile fatty acids (VFA), volatile suspended solids (VSS), pH and redox potential
27
(Eh). Once a week, a sample from inside the reactor was withdrawn for measuring VSS,
alkalinity, pH and redox potential, whereas viable cells were determined monthly. In the FBR,
the total biomass concentration was determined only in the phase V, according to the
following procedure: 5g of the colonized activated carbon was removed from within the
reactor and then submitted to sonication in an ice bath, for 5 minutes. This procedure was
performed 3 times and afterwards the carbon fraction was filtered before submitting the liquid
phase to standard VSS measurement [36].
2.2.3. Analytical methods
Sulfate concentration was determined by ionic chromatography (Metrohm) using an ASSUP10 column and conductivity detection. VFA (acetic, propionic, valeric, butyric) were analyzed
by high performance liquid chromatography, (HPLC, Shimadzu), with an ion exchange
column Aminex HPX-87H 300mm x 7.8mm (Bio-Rad). Prior to injection, samples were
filtered using 0.22µm membrane filters (Millipore). The chromatographic method was
properly validated and is detailed elsewhere [37]. Bicarbonate alkalinity (BA), VSS and COD
analysis were carried out according to the Standard Methods for Water and Wastewater [36].
Before COD determination, any sulfide present in effluent samples was stripped off by adding
a drop of HCl (35%) and flushing the sample for 10 min with N2. The solution’s pH (Hanna
HI931400) and its redox potential (Digimed) (vs an Ag/AgCl electrode) were also recorded.
Microorganisms in the liquid phase (free cells) were quantized by a three-tube most probable
number (MPN) procedure in a specific medium for SRB growth (Postgate C) [38]. Prior to the
experiments, culture tubes were degassed with pure N2, sealed and autoclaved (120 oC,
28
1.5atm, 20min). Subsequently, culture tubes plus the control were incubated for 30 days, at
35oC.
2.3. Results
The performance of both an UASB and a fluidized bed reactors treating sulfate-laden
solutions was compared under different operational conditions (figure 2.2) such as organic
and sulfate loading rates, mixing conditions and organic substrate type (only in the FBR).
Furthermore, the sulfidogenic activity was monitored through sulfate and lactate profiles and
by the volatile fatty acids produced.
2.3.1. Performance and stability of UASB and Fluidized Bed reactors.
Figure 2.3 indicates that although both reactors were started up with the same inoculum,
which were enriched with the same growth medium (modified Postgate C), the performances
of both reactors were quite distinct. A high sulfate reduction efficiency (>90%) was observed
in the FBR as soon as the adaptation phase ended (figure 2.3A), resulting in residual sulfate
concentrations below 250mg/L, already in phase I. Similar behavior was not observed in the
UASB reactor, which showed sulfate removal efficiencies between 36% and 66%, during the
phases in which the reactor operated without biomass recirculation (I-VI) (figure 2.3B).
Nevertheless, when the upflow velocity changed from 0.024m/h to 1.75m/h (phase VII)
sulfate removal increased to 89%, as depicted in figure 2.3B.
29
100
1750
A
75
1250
1000
50
750
500
25
250
0
Sulfate concentration (mg/L)
Sulfate removal efficieny (%)
1500
0
I
II
III
IV
V
VI
Phases
Sulfate removal
Residual sulfate
target
1750
1500
75
1250
1000
50
750
500
25
250
0
B
Sulfate concentration (mg/L)
Sulfate removal efficiency (%)
100
0
I
II
III
IV
V
VI
VII
Phases
Sulfate removal
Residual sulfate
target
Figure 2.3. Sulfate removal, residual and target sulfate concentrations in different phases of
the FBR (A) and UASB reactor (B) operations.
Worldwide discharge limits for sulfate in industrial wastewaters vary between 250mg/L and
500mg/L [1, 2]. Figure 2.3B indicates that the UASB reactor exhibited larger scattering in the
residual sulfate concentration as compared to the FBR (figure 2.3A). Working with a target
30
value of 250mg/L, it can be seen that during phases I, III, V and VI, the FBR resulted in
residual sulfate concentrations below that limit. Conversely, the UASB reactor was unable to
produce final sulfate concentrations below 250mg/L and the lowest average sulfate
concentration was achieved during phase VII (with recirculation), however, with a
considerable scattering (275±106mg/L), as shown in figure 2.3B.
As sulfate reduction (and alkalinity production) was constantly higher in the FBR than in the
UASB reactor, the former showed a much more stable operation. The pH fluctuated between
7.9 and 8.8, with a mean value of 8.3 (figure 2.4A) when lactate was the carbon source,
dropping to 7.5 when glycerol was the substrate. Going from phase I (COD/sulfate mass ratio
of 2.6) to phase 2 (COD/sulfate mass ratio of 1.8) there was a decrease in both alkalinity and
VFA concentrations, which is likely a consequence of both lower organic loading rates (OLR)
(figure 2.5A) and sulfate reduction yields (figure 2.3A). The FBR showed a tendency towards
stabilization during the remaining phases treating lactate (III to V), which was reflected in
values of free SRB cells above 10 9cells/mL (figure 2.6A). The large data scattering observed
in VFA figures during phase IV (figure 2.4C) can be ascribed to an increased in the sulfate
loading rate (SLR) from 5.01 ± 0.29gSO42-/(L.d) in phases I and II to 6.46 ± 0.34gSO42-/(L.d)
(figure 2.5C). In the best operational conditions (phases III and V) a VFA/alkalinity ratio
around 1 was observed. Both VFA and alkalinity were reduced during phase VI as will be
discussed subsequently in this study.
31
A
I
II
III
IV
V
B
10.0
VI
9.5
9.5
9.0
9.0
8.5
8.5
pH
pH
10.0
8.0
8.0
7.5
7.5
7.0
7.0
6.5
6.5
6.0
I
III
II
Volatile fatty acids (mgH 2C 2O 4 /L)
VII
Bicarbonate alkalinity (mgCaCO 3/L)
I
III
IV
V
II
Phases
C
VI
5 0 00
VFA and BA concentrations (mg/L)
VFA and BA concentrations (mg/L)
VI
6.0
Phases
5000
V
IV
4000
3000
2000
1000
0
Phases
D
V o latile fa tty a cid s (m g H 2 C 2 O 4 /L )
B ica rb o n ate alk alin ity (m g C aC O 3 /L )
I
II
III
IV
V
VI
V II
4 0 00
3 0 00
2 0 00
1 0 00
0
Phases
Figure 2.4. Performance parameters in different phases (according to time diagram, figure 2.2) in the FBR (A and C) and the UASB reactor (B
and D).VFA: volatile fatty acids; BA: bicarbonate alkalinity.
32
\b(Organic loading rate)
Organic removal rate
Organic loading rate
Organic removal rate
A
14
12
12
10
10
I
III
V
IV
VI
VII
9
8
8
7
7
6
6
5
5
4
4
3
3
8
8
6
6
4
4
2
2
2
2
1
1
0
0
0
I
II
III
VI
V
IV
0
Phases
Phases
Sulfate loading rate
Sulfate removal rate
I
II
III
C
IV
V
VI
8
6
6
5
5
4
4
3
3
2
2
1
1
0
0
Phase
Sulfate loading rate
Sulfate removal rate
III
I
IV
II
D
V
VI
VII
4.0
3.5
3.5
3.0
3.0
2.5
2.5
2.0
2.0
1.5
1.5
1.0
1.0
0.5
0.5
0.0
Sulfate removal rate (g/(L.d))
7
Sulfate removal rate (g/(L.d))
7
4.0
Sulfate loading rate(g/(L.d))
8
Sulfate loading rate (g/(L.d))
II
B
Organic removal rate (g/(L.d))
14
9
Organic loading rate (g/(L.d))
16
COD removal rate (g/(L.d))
Organic loading rate (g/(L.d))
16
0.0
Phases
Figure 2.5.Volumetric organic and sulfate loading rates applied and removal in the UASB reactor and the FBR. Organic loadingand removal
rates in the FBR (A) and UASB reactor (B); sulfate loading and removal rates in the FBR (C) and UASB reactor (D).
33
A
100
Lactate oxidized
by SRB
SRB population
1E10
80
70
60
1E9
50
40
1E8
30
20
10
SRB population (MPN)
Oxizided lactate (%)
90
1E7
0
I
II
III
IV
V
Phases
B
100
Lactate oxidized
75 by SRB
1E10
SRB population
1E9
50
1E8
25
1E7
SRB population (MPN)
Oxizided lactate (%)
1E11
0
I
II
III
IV
V
VI
VII
Phases
Figure 2.6. SRB population and lactate oxidized by SRB during continuous sulfate removal
in the FBR (A) and UASB reactor (B).Glycerol was utilized as substrate in phase VI during
the operation of the FBR and therefore does not appear in figure A.
34
In the UASB reactor during the phases without recirculation (I to VI) pH values increased
from 7.1 (phase I) to 7.6 (phase III), figure 2.4B, which can be related to the increase and
stabilization of the SRB population, which attained 1010free SRBcells/mL in phase III (figure
2.6B). As a result, alkalinity stabilized in the 1300-1500mg/L range (phase IV onward) as
show in figure 2.4D, which agrees with the data of Ren et al. [30], who stated that high sulfate
reduction rates (80% to 90%) required alkalinity values in this range. During phases IV and
V, the reduction in pH values (7.4 to 6.9, figure 2.4B) can be ascribed to the increase applied
in the OLR (from 4.65 ± 0.30g/(L.d) to 5.89 ± 0.48g/(L.d)), figure 2.5B, that implied in larger
VFA production (figure 2.4D). Biomass recirculation (phase VII) enabled stabilization of
both VFA and alkalinity, which resulted in higher pH values (7.5, figure 2.4B). A common
consequence of VFA build-up is the failure of anaerobic reactors, derived from a sharp drop
in pH values, which results in inhibition of the methanogenic activity [39]. However, during
sulfidogenesis there was a large production of alkalinity (figures 2.4C and 2.4D) caused by
the activity of different SRB groups (Desulfovibrio, Desulfobulbus, Desulfotomaculum,
Desulfomona), during the incomplete substrate oxidation (either lactate or glycerol) [40].
Such alkalinity enabled the pH values to be maintained in the optimum values for SRB
growth [41], without any external alkalinity requirement.
2.3.2. COD consumption and sulfate reduction yields
The profile of COD consumption (organic removal rate) can be observed in figure 2.5. In the
FBR, the substrate consumption rate varied from 4.05±0.85gCOD/(L.d), in phase II, to
7.08±1.34gCOD/(L.d), in phase V (figure 2.5A), whereas for the UASB reactor (figure 2.5B),
significantly lower removal rates were observed (from 0.88±0.52gCOD/(L.d) in phase VI to
1.50±0.52gCOD/(L.d) in phase IV) for those phases where no recirculation was performed.
35
Mixing conditions and improved of mass transfer might have accounted for such behavior
because when the recirculation was performed in the UASB reactor (phase VII), COD
consumption (figure 2.5B) increased to 1.94±0.56gCOD/(L.d), which was still lower than that
observed in the FBR. In addition, data scattering was more pronounced in the UASB reactor
as compared to the FBR, confirming the lower operational stability of the former.
Both reactors showed low overall COD removal rates (figure 2.5A and 2.5B), which can be
explained by lactate conversion to acetate by incomplete-oxidizing SRB (predominant in the
reactor biomass) and due to a lack of both acetoclastic-SRB and methanogenic activities [40].
Such behavior was confirmed by acetate accumulation in reactor effluents (figure 2.7).
Similar studies also observed that the incomplete COD removal was due to the absence of
microbial species which can metabolize acetate [42, 43].
2.3.3. Effect of the reactor configuration in sulfate removal
Mass transfer effects play an important role in the performance of high rate anaerobic reactors
[14] and this is particularly important for sulfidogenesis in the UASB reactor. It was observed
in the present work that the form whereby the biomass grew and was maintained in the reactor
affected the enrichment step and therefore competition with fermentative bacteria. This was a
result of lactate degradation by the different microbial strains identified in both reactors [40,
44]. Although lactate was not observed in the UASB reactor effluent, it took 220 days for the
SRB population to reach 10 8–109cells/mL (figure 2.6B) and thus higher sulfate reduction
efficiencies (66% - phase IV; figure 2.3B). This is consistent with other works in which a long
lag period was required to stabilize the SRB population [28, 45]. It is proposed in the present
work that the lactate not utilized for sulfate reduction was fermented because propionate was
36
observed in the reactor effluent in all phases (figure 2.7B). In addition, increased influent
COD resulted in larger propionate concentrations and such a behavior can be ascribed to poor
mixing conditions in the UASB reactor, which did not enable a faster SRB growth as it will
be discussed further in this work. Mass transfer is particularly impaired during sulfidogenesis
due to a lack of methanogenic activity and also because roughly only 50% of the total sulfide
is present as H2S, i.e. there is a small (or negligible) gas production.
Acetate
Butyrate
Acetate/butyrate conc. (mg/L)
4000
I
A
III
II
IV
V
VI
3500
3000
2500
2000
1500
1000
500
0
Phases
Acetate/propionate conc. (mg/L)
4000
Acetate
Propionate
I
II
B
III
IV
V
VI
VII
3000
2000
1000
0
Phases
Figure2.7.Acetate, butyrate and propionate profiles in the FBR (a) and the UASB reactor
(b).Details on the different phases are depicted in figure 2.2.
37
Moreover, an important limitation of UASB reactors treating sulfate is the granulation of the
biomass [46]. Several studies have shown that granular sludge formation is related to the
presence of methanogens, among other microorganisms [47, 48]. However, methanogens are
highly sensitive to high sulfide concentrations, whereas propionibacteria, which are
nucleation centers of the granules are outcompeted by SRB at high sulfate loading rates [49].
In addition, the latter also reduce granule sizes [50]. Methanogens were not identified in
neither reactors operated here, therefore granulation was not expected. Indeed, the fine and
weightless sludge observed in the UASB reactor (data not shown), was prone to washout.
Such phenomenon occurred during phase VI, when it was attempted to improve the mixing
conditions in the UASB reactor by increasing the flow rate from 0.125L/h to 0.167L/h and
thus the upflow velocity from 0.018m/h to 0.024m/h. This resulted in increased VSS
concentrations in the UASB effluent, from 80mg/L (on average) to nearly 500mgSSV/L
towards the end of the phase. Such biomass loss impaired the reactor performance with a drop
in both COD consumption (from 1.6g/(L.d) in phase V to 0.8g/(L.d) in phase VI; figure 2.5B)
and sulfate reduction (which progressively decreased from 70% to 40% during phase VI,
figure 2.3B). Omil et al. [15] also reported that increasing the upward velocity impaired
sulfidogenesis. Conversely, better mixing conditions coupled to the presence of a solid
enabled the presence of a SRB population larger than 109 free cells/mL (figure 2.6A) and
therefore much larger sulfate reductions in the FBR (above 90%), already in phase I.
Because increasing the superficial upflow velocity resulted in bacterial washout, biomass
recirculation was tested so that the upward velocity was increased to 1.75m/h, thus improving
mass transfer in the UASB reactor. Accordingly, sulfate reduction was improved to 89%
(specific activity of 1.6gSO42-/(gVSS.d)) in the UASB reactor during phase VII, as shown in
figure 2.3B. During this phase, there was higher COD consumption, lower dispersion in the
38
VFA and alkalinity values, i.e., more stable reactor operation (figures 2.4 and 2.5). This is
because, in such new configuration, no biomass washout was observed and the bacterial
population distribution throughout the UASB reactor was homogenized as shown in figure2.8.
Conversely, in the FBR the presence of immobilized biomass (larger bacterial population,
table 2.2) along with improved mass transfer (due to fluidization) enabled larger sulfate
reduction efficiencies, which reached 97% (specific sulfate reducing activity rate of 4.8gSO42/(gVSS.d)), during phase I (table 2.2). Several studies have reported efficient biomass
retention and improved mass transfer in the FBR, therefore larger reaction rates are normally
expected [51-53].
12
VSS mass (g)
10
8
6
4
2
0
III
Port A
Phases
Port B
VII
Port C
Figure2.8.Biomass profile in the UASB reactor (ports A, B and C; figure 2.1) during phase III
(no recirculation) and VII (with recirculation). Port c during phase VII was utilized for
biomass recirculation, so VSS not was determinated.
39
Table 2.2. Best parameters achieved during sulfate reduction in UASB reactor and the FBR,
treating synthetic sulfate wastewater with lactate (phases VII - UASB reactor and I – FBR) or
glycerol (phase VI - FBR).
Parameters
Chemical oxygen demand
(COD)
Organic loading rate
(OLR)
Sulfate loading rate (SLR)
COD/SO42- ratio
Volumetric COD removal
rate
Volumetric sulfate
reduction rate
Sulfate reduction
efficiency
COD removal efficiency
Overall biomass
concentration
Mean specific sulfate
reduction rate
Mean specific COD
removal rate
Unit
UASB
VII
I
VI
mg/L
5200 ± 320
5086 ± 276
4916 ± 503
gCOD/L.day
5.04 ± 0.33
12.34 ± 0.98
11.54 ± 1.19
gSO42-/L.day
g/g
2.0 ± 0.14
2.52 ± 0.21
4.82 ± 0.32
2.46 ± 0.16
4.67 ± 0.20
2.5 ± 0.3
gCOD/L.day
1.94 ± 0.56
6.25 ± 0.63
7.44 ± 1.68
gSO42-/L.day
1.60 ± 0.26
4.67 ± 0.35
4.21 ± 0.25
%
89±8
97±2
90±4
%
39±11
51±5
60±12
18.0
24.5
24.5
gSO42-/gVSS.day
0.089±0.014
0.191±0.016
0.172±0.010
gCOD/gVSS.day
0.108±0.031
0.255±0.026
0.304±0.069
gVSS/L
FBR
As shown in figure 2.3, the largest sulfate reduction efficiencies were observed for
COD/sulfate mass ratios above 2.5 (table 2.2). As the biomass concentration leveled out at
18.0gVSS/L (from phase III, onwards), the specific sulfate reduction rate was
0.084±0.014gSO42-/(gVSS.d) in the UASB reactor (figure 2.9) which is one order of
magnitude smaller than that observed in the FBR (0.191±0.016gSO42-/(gVSS.d), table 2.2),
for which, the biomass concentration was 24.5gVSS/L, considering free and attached (to
activated charcoal) cells. In addition, the absence of propionate suggested negligible
fermentative activity in the FBR treating lactate.
40
5
7
6
0.25
(gC3H5O3-/gVSS.d)
Specific sulfate reduction
rate
0.080
0.30
0.20
0.075
0.070
0.15
0.065
0.10
0.060
0.05
Specific propionate production rate
Propionate production
rate
0.085
Specific sulfate reduction rate
2(gSO4 /gVSS.d)
8
10 - 10 cell.SRB/gVSS
10 - 10 cell.SRB/gVSS
0.055
I
II
III
IV
V
VI
VII
Phases
Figure 2.9. Values of specific sulfate-reduction and propionate production rates in the UASB
reactor. Phase VI is characterized by a change in both flow rate and lactate concentration.
2.3.4. Sulfate reduction in the presence of pure glycerol as substrate
As the FBR presented the best performance during sulfate reduction in the presence of lactate
as a substrate, it was selected for further testing with a different carbon source. The study of
glycerol as carbon and electron source is justified because it is a by-product of biodiesel
industry and it is becoming widely available as biodiesel plants are commissioned throughout
the world. Therefore, glycerol is a potentially inexpensive substrate for sulfate reduction with
a smaller production of end-products [54].
Sulfate reduction with glycerol showed efficiencies of 90%, with average residual sulfate
concentrations of 200mg/L (figure 2.3A). The average specific sulfate reduction rate
(0.172±0.010gSO42-/(gVSS.d), table 2.2) was similar to that measured when lactate was the
41
only carbon source (0.191±0.016gSO42-/(gVSS.d)), whereas the average specific COD
removal rate (0.304±0.069gCOD/(gVSS.d)) was superior to the highest rate observed with
lactate (0.255±0.026gCOD/(gVSS.d)), as depicted in table 2.2. This occurred because the
presence of glycerol in the FBR changed the metabolic pathways, which affected both VFA
and alkalinity profiles in the reactor (figure 2.4C). As glycerol became the substrate (phase
VI), there was a sensible reduction in both VFA and alkalinity values. Therefore, the reactor
pH was reduced from 8.6 (phases V) to 7.5 (phase VI). Also, there was a decrease in acetate
concentration together with butyrate appearance in the effluent, suggesting glycerol
fermentation [55]. The different phenomena observed in both reactors are discussed next in
the section.
2.4. Discussion
Previous work carried out in our laboratory [40, 44] analyzed the main metabolic pathways
accounting for sulfate reduction and organic matter oxidation in both the UASB and the FBR.
This was accomplished by analyzing the relationship between microbial diversity and VFA
profile. Figure 2.10 depicts a summary of such outcomes i.e. the microorganisms identified in
the biomass during lactate oxidation along with the VFA profile in the reactor effluents.
42
Figure2.10. Main metabolic pathways developed during continuous sulfate removal in UASB
and FBR during lactate and glycerol degradation. FB - Fermenting Bacteria; SRB - Sulfate
Reducing Bacteria.
The results discussed by Bertolino et al. [40] suggested two main metabolic pathways during
lactate degradation in the UASB reactor: (i) incomplete oxidation to acetate by SRB (reaction
1); (ii) substrate fermentation to both acetate and propionate by FB such as Clostridium
(reaction 2). Conversely, incomplete lactate oxidation (reaction 1) was the predominant
metabolic pathway in the FBR [44], treating lactate. Therefore, in the present study, the
performance of both reactors can be related to the competition between SRB and FB.
2 C3H5O3- +
3 C3H5O3-
SO42→
→ 2 C2H3O2- + HS- + 2 HCO3-
C2H3O2- +
2 C3H5O2-
+
+
HCO3- +
H+
(-160.1 kJ)
H+(-169.7 kJ)
(1)
(2)
Reactions 1 and 2 explain the COD profile observed in figures 2.5A and 2.5B. Those
electrons required for sulfate reduction came from incomplete lactate oxidation (reaction 1).
In the UASB reactor, the VFA profile in the effluent (acetate and propionate build-up; figure
2.7B) suggested that a significant fraction of the influent COD (CODlactate) was converted to
acetate (CODacetate) by SRB (during sulfate reduction) and the remaining COD was fermented
43
to acetate (CODacetate)and propionate (CODpropionate) by Clostridium spp [40]. As the organic
loading rate increased from 3.55±0.25g/(L.d) (phase III) to 5.89±0.48g/(L.d) (phases V), there
was an increase in propionate concentrations (from 0.043±0.018g/(gVSS.d) to 0.157±0.019
g/(gVSS.d)) (stronger fermentative activity)and reduced sulfidogenic activity (from
0.077gSO42-/(gVSS.d) to 0.057gSO42-/(gVSS.d), respectively), figure 2.9. Such behavior is
consistent with the work of Oyekola et al. [56], who also observed fully lactate conversion
and an increase in propionate concentration as the organic loading rate increased. Biomass
recirculation (phase VII – UASB reactor) led to an increase in the specific sulfate-reduction
rate and also a decrease in propionate production (figure 2.9), suggesting a lower fermentative
activity.
Furthermore, assuming incomplete lactate oxidation to acetate by SRB (reaction 1), only
360mgCOD/g-lactate would be oxidized (reactions 3 and 4) during sulfate reduction.
Therefore, for the reduction of 2gSO42-/L, nearly 1.33gCOD/L (reaction 4) would be
consumed, which correspond to 3.71g/L of lactate (4.0gCOD/L), i.e, lactate would not limit
sulfate reduction for COD/sulfate ratios above 2. Therefore, supposing that SRB utilized only
lactate during phases II, IV, V and VI there would not be organic substrate limitation in the
UASB reactor, even though sulfate reduction efficiencies were low - from 49% (phase II OLR = 4.87±0.30g/(L.d)) to 66% (phase IV, OLR = 4.65±0.30g/(L.d)). The poor sulfate
reduction performance can be explained by the poor mixing conditions in the UASB reactor
(figure 2.8), as stated, which enhanced the fermentative activity.
C3H5O3- +O2→ C2H3O2- + CO2+ H2O
S2-
+ 2O2 ⇆ SO42-
(3)
(4)
44
From the amount of reduced sulfate (figure 2.5), the fraction of lactate utilized by SRB can be
estimated according to reaction 1. During phases I and II (UASB reactor), a low SRB
population
(figure
2.3)
(1.01±0.36gCOD/(L.d))
implied
was
that
utilized
only
for
38%-40%
sulfate
of
reduction
the
oxidized
lactate
(0.68±0.18gSO42-/(L.d)-
0.93±0.23gSO42-/(L.d); figure 2.5). The remaining lactate was then degraded by FB because
the effluent lactate concentration was always negligible [40]. This is consistent with the fact
that lactate fermenters have a higher growth rate and a lower affinity for lactate [56], as
suggested by Zellner et al. [57]. The authors determined µ maxand Ks for Desulfovibrio as
0.25h-1 and 1.5mmol/L, respectively, whereas for Clostridium sp. the same parameters were
0.7h-1 and 2.5mmol/L, respectively. As a result, propionate concentrations in the UASB
effluent were particularly high in those phases in which high organic loadings were applied
(IV and V).
Similar behavior was not observed in the FBR treating lactate, in which fermentative activity
was negligible (because propionate was absent [44]) and the influent COD (CODlactate) was
converted entirely to acetate (CODacetate), figure 2.7A, according to reaction 1. Sulfate
reduction was lower only in those phases where the COD/sulfate ratio was below 2 (II and
IV). For instance, during phases I, 97±2% sulfate reduction was observed for an OLR of
12.34±0.98gCOD/(L.d) (COD/sulfate mass ratio > 2.5) as compared to 78±10%, when the
OLR was 8.7±0.63gCOD/(L.d) (COD/sulfate mass ratio of 1.8) in phase II.
A metabolic pathway for the oxidation of glycerol during sulfate reduction was hypothesized
by Dinkel et al [29], which is presented in reaction 5. Reaction 5 predicts that alkalinity
should be lower than that produced during lactate degradation (reaction 1), explaining the
experimental results achieved in the FBR (figure 2.4C, phase VI). The results herein
45
presented also show that during glycerol degradation, acetate (CODacetate) was still the main
reaction product, but a small fraction of the effluent COD was due to the presence of
150mg/L of butyrate (CODbutyrate) (figure 2.7A). Butyrate presence in the reactor suggested
fermentation by Clostridium sp. (identified in the biomass, [44]) because such
microorganisms were shown to ferment glycerol, producing the observed acetate and butyrate
[58]. Therefore, assuming that acetate was not oxidized by either MPM or SRB [40],the
stoichiometry of equations 1 and 5 suggested that 50% t of produced acetate (809±143mg/L)
was due to glycerol oxidation by SRB (particularly Desulfovibrio spp) and the other 50% can
be related to glycerol fermentation by Clostridium ssp. Such outcomes suggest that glycerol is
not as easily degradable as lactate. Indeed, the maximum specific growth rate of SRB on
glycerol-base medium was reported as 0.056h-1[29], which is one order de magnitude lower
than that reported for SRB growth on lactate [57]. It seems that glycerol needs to be first
degraded to an intermediary product before being utilized by SRB [44].
C3 H8O3+ 1.25 SO42- →
0.5 C2H3O2-+ 1.5 H2CO3 + 0.5 HCO3- + 1.25 HS- + 0.75OH- + 0.25H2O
(-424.5 kJ)
(5)
Acetate build-up is reported as a drawback in high-rate sulfate-reducing reactors [4, 10, 11]
because the amount of residual COD in the reactor effluent requires downstream treatment. In
this regard, the present study has demonstrated that sulfate reduction in the presence of
glycerol as organic substrate produced a smaller residual COD (1700mg/L) than that observed
with lactate (2500mg/L C2 H3O2-) at the same COD/sulfate mass ratio (2.5).Such values are
even smaller than those produced (2660mg/L C2 H3O2-) when ethanol (utilized in industrial
scale sulfate-reducing plants) was applied as carbon and electron source [11]. The produced
sulfide can be separated from acetate by either precipitation with transition metals (Fe, Cu,
46
Ni) [22] or stripping by an inert gas (N2 or CO2) as proposed by Marre et al. [59], or even
oxidized to elemental sulfur (by Fe3+ or NO3-). After H2S removal, acetate can be degraded
either aerobically or anaerobically depending on the process configuration and feed water
quality. Overall, as a by-product of the emerging biodiesel industry, crude glycerol may be
foreseen as a cost-effective alternative to lactate and ethanol for sulfate reduction. Future
work will focus on the application of crude glycerol for sulfate removal.
2.5. Conclusions
Mixing conditions plays a key role during sulfidogenesis. Lactate fermentation by
Clostridium spp. was an important metabolic pathway in an bench scale UASB reactor
treating 2.0g/(L.d) sulfate, without biomass recirculation (poor mixing conditions). The
maximum volumetric sulfate reduction rate was 1.3gSO42-/(L.d)) (66% removal), whereas
fermentation resulted in a high propionate production rate (3.91g/(L.d)). An increase in the
upflow velocity, from 0.125m/h to 1.75m/h, due to recirculation improved biomass
distribution in the reactor and thus the sulfate removal rate to 1.6gSO42-/(L.d) (89% removal),
but decreased the propionate production rate to 0.88g/(L.d). Therefore, improved mixing
conditions in the UASB reactor enhanced both substrate degradation and sulfate reduction, as
opposed to substrate fermentation. In the fluidized bed reactor, better mass transfer conditions
enabled the predominance of sulfate reducing activity by incomplete-oxidizing SRB. When
sulfate was not limiting (COD/sulfate mass ratios higher than 2), the sulfate removal rate
varied between 4.7g/(L.d) and 5.1g/(L.d), which corresponds to sulfate removal efficiencies
higher than 95%. The FBR was able to utilize pure glycerol as carbon and electron source,
producing sulfate reduction rates (0.172±0.010gSO42-/(gSSV.d) similar to those observed with
lactate (0.191±0.016gSO42-/(gSSV.d)). As a by-product of the biodiesel industry, glycerol can
47
be a cost-effective option for sulfate reduction leading to lower acetate concentrations
(1700mg/L) when compared to lactate oxidation (2500mg/L).
2.6. Acknowledgements
The financial support from the funding agencies FINEP, FAPEMIG, CNPq and CAPES is
gratefully appreciated. The scholarships to S. M. Bertolino, S. F. Aquino and V. A. Leão are
especially acknowledged.
2.7. References
[1] INAP, Treatment of sulphate in mine effluentes, in, International network for acid
prevention, 2003, pp. 129.
[2] WHO, Guidelines for drinking-water quality, in, Genebra, 2011, pp. 564.
[3] M.M.M. Gonçalves, S.G.F. Leite, G.L. Sant'Anna Jr, The bioactivation procedure for
increasing the sulphate-reducing bacteria in a UASB reactor, Brazilian Journal of Chemical
Engineering, 22 (2005) 565-571.
[4] A. Kaksonen, P. Franzmann, J. Puhakka, Performance and Ethanol Oxidation Kinetics of
a Sulfate-Reducing Fluidized-Bed Reactor Treating Acidic Metal-Containing Wastewater,
Biodegradation, 14 (2003) 207-217.
[5] K. Kosinska, T. Miskiewicz, Performance of an anaerobic bioreactor with biomass
recycling, continuously removing COD and sulphate from industrial wastes, Bioresource
Technology, 100 (2009) 86-90.
[6] S.I.C. Lopes, M.I. Capela, P.N.L. Lens, Sulfate reduction during the acidification of
sucrose at pH 5 under thermophilic (55 °C) conditions. II: Effect of sulfide and COD/ SO 4 2
- ratio, Bioresource Technology, 101 (2010) 4278-4284.
[7] J. Poinapen, G. Ekama, M. Wentzel, Biological sulphate reduction with primary sewage
sludge in an upflow anaerobic sludge bed (UASB) reactor - Part 2: Modification of simple
wet chemistry analytical procedures to achieve COD and S mass balances, Water SA, 35
(2009) 535-542.
48
[8] J. Shayegan, F. Ghavipanjeh, P. Mirjafari, The effect of influent COD and upward flow
velocity on the behaviour of sulphate-reducing bacteria, Process Biochemistry, 40 (2005)
2305-2310.
[9] M.V.G. Vallero, R.H.M. Treviño, P.L. Paulo, G. Lettinga, P.N.L. Lens, Effect of sulfate
on methanol degradation in thermophilic (55 °C) methanogenic UASB reactors, Enzyme and
Microbial Technology, 32 (2003) 676-687.
[10] L.B. Celis-García, E. Razo-Flores, O. Monroy, Performance of a down-flow fluidized
bed reactor under sulfate reduction conditions using volatile fatty acids as electron donors,
Biotechnology and Bioengineering, 97 (2007) 771-779.
[11] S. Nagpal, S. Chuichulcherm, L. Peeva, A. Livingston, Microbial sulfate reduction in a
liquid–solid fluidized bed reactor, Biotechnology and Bioengineering, 70 (2000) 370-380.
[12] L.M. Nevatalo, A.E. Mäkinen, A.H. Kaksonen, J.A. Puhakka, Biological hydrogen
sulfide production in an ethanol lactate fed fluidized-bed bioreactor, Bioresource Technology,
101 (2010) 276-284.
[13] E. Sahinkaya, B. Özkaya, A.H. Kaksonen, J.A. Puhakka, Sulfidogenic fluidized-bed
treatment of metal-containing wastewater at low and high temperatures, Biotechnology and
Bioengineering, 96 (2007) 1064-1072.
[14] M.T. Kato, J.A. Field, P. Versteeg, G. Lettinga, Feasibility of expanded granular sludge
bed reactors for the anaerobic treatment of low-strength soluble wastewaters, Biotechnology
and Bioengineering, 44 (1994) 469-479.
[15] F. Omil, P. Lens, L.W. Hulshoff Pol, G. Lettinga, Effect of upward velocity and sulphide
concentration on volatile fatty acid degradation in a sulphidogenic granular sludge, Process
Biochemistry 31 (1996) 699-710., (1996).
[16] P.H. Nielsen, Biofilm Dynamics and Kinetics during High-Rate Sulfate Reduction under
Anaerobic Conditions, Applied and Environmental Microbiology, 53 (1987) 27-32.
[17] S.N. Medírcio, Redução do teor de sulfato em águas pela utilização de bactérias
redutoras de sulfato (Sulfate removal from wastewaters with Sulfate Reducing Bacteria), in,
Universidade Federal de Ouro Preto, Ouro Preto, MG, Brazil, 2004, pp. 78.
[18] A.H. Kaksonen, J.J. Plumb, P.D. Franzmann, J.A. Puhakka, Simple organic electron
donors support diverse sulfate-reducing communities in fluidized-bed reactors treating acidic
metal- and sulfate-containing wastewater, FEMS Microbiology Ecology, 47 (2004) 279-289.
[19] W. Liamleam, A.P. Annachhatre, Electron donors for biological sulfate reduction,
Biotechnology Advances, 25 (2007) 452-463.
49
[20] A. Velasco, M. Ramírez, T. Volke-Sepúlveda, A. González-Sánchez, S. Revah,
Evaluation of feed COD/sulfate ratio as a control criterion for the biological hydrogen sulfide
production and lead precipitation, Journal of Hazardous Materials, 151 (2008) 407-413.
[21] S.V. Mohan, Prasad, K.K., Rao, N.C., Bhaskar, Y.v., Babu, V.L., Rajagopal, D. and
Sarma, P.N., Biological treatment of low-biodegradable composite chemical wastewater using
upflow anaerobic sludge blanket (UASB) reactor, Journal of scientific and industrial research,
64 (2005) 771-777.
[22] J. Cao, G. Zhang, Z. Mao, Z. Fang, C. Yang, Precipitation of valuable metals from
bioleaching solution by biogenic sulfides, Minerals Engineering, 22 (2009) 289-295.
[23] S. Dar, R. Kleerebezem, A. Stams, J. Kuenen, G. Muyzer, Competition and coexistence
of sulfate-reducing bacteria, acetogens and methanogens in a lab-scale anaerobic bioreactor as
affected by changing substrate to sulfate ratio, Applied Microbiology and Biotechnology, 78
(2008) 1045-1055.
[24] Y. Zhao, N. Ren, A. Wang, Contributions of fermentative acidogenic bacteria and
sulfate-reducing bacteria to lactate degradation and sulfate reduction, Chemosphere, 72
(2008) 233-242.
[25] A.S. Sheoran, V. Sheoran, R.P. Choudhary, Bioremediation of acid-rock drainage by
sulphate-reducing prokaryotes: A review, Minerals Engineering, 23 (2010) 1073-1100.
[26] S.K. Bhattacharya, V. Uberoi, M.M. Dronamraju, Interaction between acetate fed sulfate
reducers and methanogens, Water Research, 30 (1996) 2239-2246.
[27] H. Harada, S. Uemura, K. Momonoi, Interaction between sulfate-reducing bacteria and
methane-producing bacteria in UASB reactors fed with low strength wastes containing
different levels of sulfate, Water Research, 28 (1994) 355-367.
[28] F. Omil, P. Lens, A. Visser, L.W. Hulshoff Pol, G. Lettinga, Long-term competition
between sulfate reducing and methanogenic bacteria in UASB reactors treating volatile fatty
acids, Biotechnology and Bioengineering, 57 (1998) 676-685.
[29] V. Dinkel, F. Frechen, A. Dinkel, Y. Smirnov, S. Kalyuzhnyi, Kinetics of anaerobic
biodegradation of glycerol by sulfate-reducing bacteria, Applied Biochemistry and
Microbiology, 46 (2010) 712-718.
[30] N.-Q. Ren, H. Chua, S.-Y.Chan, Y.-F. Tsang, N. Sin, Effects of COD/SO42- Ratios on
an Acidogenic Sulfate-Reducing Reactor, Industrial & Engineering Chemistry Research, 46
(2007) 1661-1666.
50
[31] J.A. Alvarez, L. Otero, J.M. Lema, A methodology for optimising feed composition for
anaerobic co-digestion of agro-industrial waste., Bioresource Technology, 101 (2010) 11531158.
[32] M.S. Fountoulakis, K. Czaczyk, Enhanced methane and hydrogen production from
municipal solid waste and agro-industrial by-products co-digested with crude glycerol.,
Bioresource Technology, 100 (2009) 3043-3047.
[33] J.A.S. Lopéz, M.A.M. Santos, A.F.C. Pérez, Anaerobic digestion of glycerol derived
from biodiesel manufacturing., Bioresource Technology, 100 (2009) 5609-5615.
[34] Y. Yang, K. Tsukahara, S. Sawayama, Biodegradation and methane production from
glycerol-contaning synthetic waste with fixed-bed bioreactor under mesophilic and
thermophilic anaerobic conditions Process Biochemistry, 43 (2008) 362-367.
[35] A.I. Qatibi, Fermentation du lactate, du glycerol et des diols par les bactéris sulfatoréductrices du genere Desulfovibrio, in: Microbiology, University of Provence, France, 1990,
pp. 165.
[36] APHA, Standard Methods for the examination of water and wastewater, Washington,
DC, 2005.
[37] P.L. Mesquita, S.F. Aquino, A.L.P. Xavier, J.C.C.d. Silva, R.C.F. Afonso, S.Q. Silva,
Soluble microbial product (SMP) characterization in bench-scale aerobic and anaerobic
CSTRs under different operational conditions, Brazilian Journal of Chemical Engineering, 27
(2010) 101-111.
[38] J.R. Postgate, Versatile medium for the enumeration of sulfate-reducing bacteria.,
Applied Microbiology, 11 (1963) 265-267.
[39] S.F. Aquino, C.A. Chernicaro, Build up of volatile fatty acids (VFA) in anaerobic
reactors under stress conditions: causes and control strategies, Engenharia Sanitaria e
Ambiental 10 (2005).
[40] S.M. Bertolino, I.C.B. Rodrigues, R. Guerra-Sá, S.F. Aquino, V.A. Leão, Implications of
volatile fatty acid profile on the metabolic pathway during continuous sulfate reduction,
Journal of Environmental Management, 103 (2012) 15-23.
[41] L.L. Barton, Sulfate-reducing bacteria, Plenum Press, New York, 1995.
[42] L. Celis-Garcia, D. Villa-Gómez, A. Alpuche-Solís, B. Ortega-Morales, E. Razo-Flores,
Characterization of sulfate-reducing bacteria dominated surface communities during start-up
of a down-flow fluidized bed reactor, Journal of Industrial Microbiology & Biotechnology, 36
(2009) 111-121.
51
[43] T. Yamaguchi, H. Harada, T. Hisano, S. Yamazaki, I.C. Tseng, Process behavior of
UASB reactor treating a wastewater containing high strength sulfate, Water Research, 33
(1999) 3182-3190.
[44] S.M. Bertolino, L.A. Melgaço, S.F. Aquino, V.A. Leão, Glycerol as a single electron
donor for sulfate reduction on fluidized bed reactors, Water Research, Submited Manuscript
(2012).
[45] S. Beaulieu, G.J. Zagury, R. Samson, Bioactivation and bioaugmentation of a passive
reactor for acid mine drainage treatment, in: R.K. Singhal, A.K. Mehrotra (Eds.)
Environmental Issues and Management of Waste in Energy and Mineral Production,
Rotterdam, The Netherlands, 2000, pp. 533-537.
[46] R.E. Speece, Anaerobic biotechnology for industrial wastewaster treatment,
Environmental Science Technology 17 (1983) 416 - 427.
[47] J.E. Schmidt, B.K. Ahring, Granular sludge formation in upflow anaerobic sludge
blanket (UASB) reactors, Biotechnology and Bioengineering, 49 (1996) 229-246.
[48] Y. Liu, H.-L.Xu, S.-F.Yang, J.-H. Tay, Mechanisms and models for anaerobic
granulation in upflow anaerobic sludge blanket reactor, Water Research, 37 (2003) 661-673.
[49] A. Overmeire, P. Lens, W. Verstraete, Mass Transfer Limitation of Sulfate in
Methanogenic Aggregates, Biotechnology and Bioengineering, 44 (1994) 387-391.
[50] A. Visser, P.A. Alphenaar, Y. Gao, G. van Rossum, G. Lettinga, Granulation and
immobilisation of methanogenic and sulfate-reducing bacteria in high-rate anaerobic reactors.,
Journal Applied Microbiology and Biotechnology, 40 (1993) 575-581.
[51] V.S. Steed, M.T. Suidan, M. Gupta, T. Miyahara, C.M. Acheson, G.D. Sayles,
Development of a sulfate-reducing biological process to remove heavy metals from acid mine
drainage, Water Environment Research, 72 (2000) 530-535.
[52] P. Marín, D. Alkalay, L. Guerrero, R. Chamy, Desing and startup of an anaerobic
fluidized bed reactor, Water Science Technology, 40 (1999) 63-70.
[53] W.K. Shieh, J.D. Keenan, Fluidized bed biofilm reactor for wastewater treatment, Adv.
Biochemistry Eng, Biotechonology, 33 (1986) 131-169.
[54] N. Kolesárová, M. Hutnan, I. Bodík, V. Spalková, Utilization of Biodiesel By-Products
for Biogas Production, Journal of Biomedicine and Biotechnology, 2011 (2011) 1-15.
52
[55] K. Leja, K. Czaczyk, K. Myszka, Biotecnhological synthesis of 1,3-propanediol using
Clostridium ssp., African Journal of Biotechnology 10 (2011) 11093-11101.
[56] O.O. Oyekola, R.P. van Hille, S.T.L. Harrison, Kinetic analysis of biological sulphate
reduction using lactate as carbon source and electron donor: Effect of sulphate concentration,
Chemical Engineering Science, 65 (2010) 4771-4781.
[57] G. Zellner, F. Neudorfer, H. Diekmann, Degradation of Lactate by an Anaerobic Mixed
Culture in a Fluidized-Bed Reactor, Water Research, 28 (1994) 1337-1340.
[58] A. Drozdzynska, K. Leja, K. Czaczyk, Biotechnological production of 1,3-propanediol
from crude glycerol, Journal of Biotechonology 92 (2011) 92-100.
[59] J. Maree, G. Strobos, H. Greben, E. Netshidaulu, E. Steyn, A. Christie, P. nther, F.
Waanders, Treatment of Acid Leachate from Coal Discard using Calcium Carbonate and
Biological Sulphate Removal, Mine Water and the Environment, 23 (2004) 144-151.
53
CAPÍTULO 3
IMPLICATIONS OF VOLATILE FATTY ACID PROFILE ON THE METABOLIC
PATHWAY DURING CONTINUOUS SULFATE REDUCTION
Abstract
Volatile fatty acid (VFA) profile is an important parameter in anaerobic reactors because it
enables the assessment of metabolic pathways. Volatile fatty acids were monitored during
sulfate reduction in a UASB (upflow anaerobic sludge blanket) reactor treating 2g/L sulfate
concentration and with the organic loading increasing from 3.5 kgCOD/m3.d to 5.9
kgCOD/m3.d, for a 1-day residence time. In the absence of recirculation, the best outcome
(65% reduction) was noticed with the lowest organic loading (3.55 kg/m3.d). When
recirculation was applied, sulfate reduction yields increased to 89%, corresponding to a
sulfate removal rate of 1.94 kgSO42-/m3.d. The reactor performance was discussed in relation
to microbial diversity and metabolic pathways. At high organic loading, two metabolic
pathways accounted for lactate degradation: (i) lactate is oxidized to acetate and carbon
dioxide by the incomplete-oxidizer SRB (sulfate-reducing bacteria) Desulfomonas,
Desulfovibrio, Desulfolobus, Desulfobulbus and Desulfotomaculum spp.; (ii) lactate is
converted to acetate by fermenting bacteria such as Clostridium sp. High propionate
concentrations imply that there is low sulfate reduction efficiency.
Key-words: UASB reactor, sulfate reduction, lactate, propionate, fermentation, metabolic
pathways.
54
3.1. Introduction
Sulfate is always present in effluents from the chemical, metallurgical and pharmaceutical
industries because of the widespread use of sulfuric acid. The former can also be produced
during natural oxidation of sulfide minerals. Although sulfate is considered a low-risk
substance when compared to dissolved metals and acidity, regulatory agencies are becoming
increasingly concerned about high sulfate levels in effluents and stricter standards are being
imposed or expected in the near future (INAP, 2003; USEPA, 2009; WHO, 2011).
The treatment of sulfate containing wastewaters comprises both biological and chemical
processes. Chemical processes can sometimes be expensive and produce a high volume of
sludge. Biological treatment uses sulfate-reducing bacteria (SRB), which is present in many
anaerobic wastewater treatment systems and this route has been extensively studied because
sulfate and metal concentrations can be reduced to very low levels, sludge production is
minimal and the precipitated metal sulfides can be sold; reducing operational costs.
Conversely, biological sulfate reduction economics relies heavily on the carbon source and
the process is also influenced by temperature; and thereby drainage heating in moderate
climates may be required.
SRB utilize easily degradable organic compounds, including lactate, which can be oxidized
by different species (Barton, 1995; Liamleam and Annachhatre, 2007). When lactate is
biologically degraded, volatile fatty acids such as pyruvate, acetate, propionate and butyrate
can be produced (García, 1982). These compounds can also be degraded by either SRB or
other microorganisms. As such the VFA profile can be related to the different phenomena
occurring in anaerobic reactors (acidogenesis, sulfidogenesis and methanogenesis) (Aquino
55
and Chernicaro, 2005), supporting a discussion on the metabolic pathways accounting for
sulfate reduction and organic matter oxidation.
The upflow anaerobic sludge blanket (UASB) is an established anaerobic reactor. Its
advantages as compared to other anaerobic reactors include low investment and energy costs,
and short hydraulic retention time with no support medium required (Lettinga and Hulshoff
Pol, 1991) i.e. it has an simple and inexpensive design. It has some disadvantages regarding
sulfate reduction: namely (i) mixing is provided solely by the flow rate because gas
production is low or inexistent and (ii) SRB does not granulate as well as methanogenic
microorganisms. Nevertheless, this reactor has been investigated for sulfate reduction in many
studies where different carbon sources (Gonçalves et al., 2007; Harada et al., 1994; Lopes et
al., 2010; Lopes et al., 2007; Poinapen et al., 2009a; Vallero et al., 2003) as well as reactor
operational configurations were studied (Kaksonen et al., 2004; Mohan, 2005; Shayegan et
al., 2005).
Two types of studies can be cited regarding anaerobic sulfate reduction. The first type are
works where lab scale UASB reactors are investigated and parameters such as substrate type,
COD/sulfate ratio and sulfate loads are studied (table 1.1)(Gonçalves et al., 2007; Kaksonen
et al., 2003a; Kosinska and Miskiewicz, 2009; Lopes et al., 2007; Mohan, 2005; Poinapen et
al., 2009b; Shayegan et al., 2005; Vallero et al., 2003). Usually, lactate is not the chosen
substrate (due to economic constraints) and no detailed discussion on the metabolic pathways
is carried out. The second group of studies, usually performed in chemostats (Dar et al., 2008;
Zhao et al., 2008), investigates metabolic pathways with the support of molecular biology. In
this second group, lactate is the preferred carbon and electron source because it enables the
growth of different SRB strains, as stated. Therefore, this work is an attempt to apply these
56
two approaches to discuss comprehensively the performance of a UASB reactor treating high
sulfate loads. Lactate was chosen as the carbon source as it enables a deeper understanding of
the different phases occurring in anaerobic reactors.
3.2. Experimental
3.2.1. Microorganisms and growth medium
The inoculum used in this study was obtained from a granular sludge collected from an
UASB reactor treating domestic wastewater and enriched in a modified Postgate C medium
so that a 5-liter sample was produced. The enrichment medium was comprised of: 0.5g/L
KH2PO4; 1.0g/L NH4Cl; 0.06g/L MgSO4.7H2O; 0.1g/L FeSO4.7H2O; 0.25g/L yeast extract;
2.96g/L Na2SO4; and 3.76g/L lactate. Afterwards, the inoculum, containing 7.6gVSS (volatile
suspended solids) was transferred to the reactor and pumping of the growth medium
containing sulfate was started in a semi-batch mode. This process involved 24 hours of
growth medium pumping and 24-hour rest periods so that adaptation to the new reactor
(UASB) was accomplished. As soon as the whole reactor was filled with the medium and
biomass, the continuous operation was started.
3.2.2. Anaerobic reactor and operational methods
Figure 3.1 shows a schematic diagram of the of lab-scale UASB reactor (1). The total volume
was 3.0L and it was placed inside a fume hood in a temperature-controlled room whereby the
temperature was maintained at 25±2ºC. Three sampling ports (a, b and c), a gas outlet (5),
completed the reactor. A peristaltic pump (3) pumped the solution (growth medium) from the
57
feed tank (2) into the reactor (1). For recirculation, a second pump (4) was added and the
solution from port c was recycled. The effluent was collected in a second tank (6).
Figure 3.1. Schematic diagram of the UASB reactor for sulfate reduction.
This reactor operated during 580 days, at an hydraulic retention time (HRT) of 24h and was
fed with a synthetic effluent (modified Postage C medium, section 3.2.1) containing lactate as
the only carbon and electron source. The organic load varied according to the sulfidogenic
performance shown by the reactor, starting at a COD/Sulfate mass ratio of 1.8 (table 3.1).
Phases I and II represented the SRB enrichment period, whereas phases III to V were run with
increasing organic load for a constant sulfate concentration (2.0 g/L). Phase VI is
characterized by effluent recirculation (rate = 93), for an organic load rate (OLR) set at 5
kgCOD/m3.d (COD/sulfate = 2.5). The reactor effluent was analyzed twice a week for total
58
(data not shown) and filtered chemical oxygen demand (COD), sulfate, alkalinity, VFA,
volatile suspended solids (VSS), pH and redox potential (Eh). Once a week, a sample from
inside the reactor was withdrawn for measuring VSS, alkalinity, pH and redox potential,
whereas viable cells were determined monthly.
Table 3.1.Operational parameters during sulfate reduction in the UASB reactor.Hydraulic
retention time: 24 hours, 25ºC.
Phases
Days
OLRa
(kg/m3.d)
42-114
3.48 ± 0.33
I
115-216
4.87 ± 0.30
II
217-317
3.55 ± 0.25
III
318-368
4.65 ± 0.30
IV
369-445
5.89 ± 0.48
V
499-578
5.04 ± 0.33
VI
a
OLR: organic loading rate.
b
COD: chemical oxygen demand.
CODb
(mg/L)
SO42(mg/L)
COD/SO42Ratio
3512 ± 325
1967 ± 189
1.80 ± 0.20
5000 ± 280
1964 ± 101
2.55 ± 0.20
3645 ± 304
2200 ± 197
1.67 ± 0.18
4790 ± 396
2037 ± 226
2.39 ± 0.33
6040 ± 411
1944 ± 97
3.12 ± 0.28
5200 ± 332
2046 ± 140
2.52 ± 0.21
Sulfate concentration was determined by ionic chromatography (Metrohm) using an ASSUP10 column and conductivity detection. VFA (acetic, propionic, valeric, butyric) were analyzed
by high performance liquid chromatography, (HPLC, Shimadzu), with an ion exchange
column Aminex HPX-87H 300 mm x 7.8 mm (Bio-Rad). Prior to injection, samples were
filtered using 0.22 µm membrane filters (Millipore, Corp.). Bicarbonate alkalinity (BA) was
assayed by titration with 0.1 M sulfuric acid solution to pH 4.5; VSS and COD according to
the Standard Methods for Water and Wastewater (APHA, 2005). Before COD determination,
any sulfide present in effluent samples was removed by adding a drop of HCl (35%) and
flushing the sample for 10 min with N2. The solution’s pH (Hanna HI931400) and its redox
potential (Digimed) (vs an Ag/AgCl electrode) were also recorded.
59
Microorganismswere enumerated by a three-tube most probable number (MPN) procedure in
a specific medium for SRB (Postgate C) (Postgate, 1963). Prior to the experiments, culture
tubes were degassed with pure N2, sealed and autoclaved (120oC, 1.5atm, 20min).
Afterwards, culture tubes and the control tube were incubated for 30 days, at 35 oC.
16S rRNA gene sequences were utilized to study bacterial phylogeny and taxonomy present
in the sludge inoculum and in the reactor during phases I (enrichment) and IV (OLR). Briefly,
the 16S rRNA amplicons of all samples were cloned into pGEMT-Easy vector and then
sequenced in an ABI 3100 automated sequencer (Applied Biosystem), using a dye terminator
kit. The sequences were then used for phylogenic analysis. Experimental details were
described in Rampinelli et al. (2008).
3.3. Results and discussion
3.3.1. Reactor start-up and biomass
Total biomass and SRB population were followed by the biomass weight (as VSS) and the
MPN technique, respectively. The results, depicted in figure 3.2, were assessed at the end of
each phase and show a 1000-time increase in the SRB population from phase I (5.3x106
cells/mL) to phase III (9.5 x 109 cells/mL). This linear increase suggests that up to phase III,
the SRB population had not reached its maximum value. This is consistent with other works
in which a long period was required to stabilize the SRB population (Beaulieu et al., 2000;
Omil et al., 1998). Similar populations were determined by Mizuno et al. (1998), studying
sulfate reduction with sucrose in batch reactors. After phase III, the VSS values stabilized in
the range 7.62 to 13.61 gVSS, without large variations. In addition, the changes in the SRB
60
population, after phase III, reflect the changes in the COD/sulfate mass ratio as will be
16
1E11
14
1E10
12
10
1E9
8
1E8
6
4
Biomass mass (gVSS)
SRB population (cells/mL, MPN)
discussed throughout this work.
1E7
2
0
I
II
III
IV
Phases
volatile suspend solids mass
V
VI
SRB cell number
Figure 3.2. Evolution of biomass monitored in the UASB reactor: (I) OLR = 3.48 kg/m3.d;
(II) OLR = 4.87 kg/m3.d; (III) OLR = 3.55 kg/m 3.d; (IV) OLR = 4.65 kg/m3.d; (V) OLR =
5.89 kg/m3.d; (VI) OLR = 5.04 kg/m3.d.
Microbial diversity was determined in the sludge inoculum and in the reactor during phases I
(enrichment) and IV (OLR). The enrichment procedure successfully resulted in a diverse SRB
population as shown in table 3.2, while inhibiting the growth of methanogens, as the latter
were not identified in the inoculum. Nevertheless, microorganisms producing methane
(MPM) were identified in all samples taken from the reactor (phases I and IV). This microbial
diversity was expected, due to the inoculum origin (domestic sewage treating reactor). In
addition, enrichment with Postgate C medium induced, as expected, the growth of incomplete
oxidizers – those microorganisms which oxidize lactate to acetate, such as the Desulfomonas,
61
Desulfovibrio, Desulfolobus, Desulfobulbus and Desulfotomaculum genera. It must be pointed
out that although methanogens were detected in phases I and IV, their growth is more
inhibited by the presence of sulfide (especially H2S) when compared to SRB (O'Flaherty et
al., 1998). Therefore, their population was expected to decrease, as the SRB predominated in
the reactor and sulfate reduction increased (Bhattacharya et al., 1996; Briones et al., 2009;
Omil et al., 1998).
Table 3.2. Microorganisms identified by molecular biology techniques in the inoculum and
different phases during UASB reactor operation.
Microorganism
1. Desulfomonas pigra (SF192152) (IO)
2. Desulfovibrio desulfuricans subsp. Desulfuricans
str. ATCC 27774 (IO)
3. Desulfolobus sp. (IO)
4. Desulfovibrio vulgaris (IO)
5. Uncultured Desulfovibrio sp. Clone A37bac 16S
ribosomal(IO)
6. Desulfobulbus sp. (EF442937) (PO)
7. Desulfobacter halotolerans DSM 11383
(NR026439) (AO)
8. Uncultured Desulfotomaculum sp. Clone BNB488 (FJ898345) (IO)
9. Methanogens
10. Clostridium sp.
Start-up
inoculum
+
UASB
Phase I and IV
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
*
+
+
Similarity**
%
100
100
100
100
100
100
100
100
100
100
IO - Lactateincomplete – oxidizer SRB to acetate.
PO - Propionate – oxidizer SRB to acetate.
AO- Acetate – oxidizer SRB.
* -Not analyzed
** - DNA sequence deposited in gene Bank
62
3.3.2. Reactor performance
Parameters
The process stability in the UASB reactor was monitored by pH, redox potential (Ag/AgCl),
VFA concentration and alkalinity (figure 3.3), whereas, the performance of sulfate reduction
was investigated by the amount of sulfate and COD removed as well as the VFA profile.
(A)
(B)
I
IV
III
II
V
VI
-220
7.5
-240
-260
pH
-280
pH
6.5
-300
-320
-340
6.0
-360
Eh
-380
-400
4.5
-420
-440
4.0
-460
50
100
150
200
250
300
350
400
450
500
550
600
Eh (mV)
7.0
Volatile fatty Acids (mgHAc/L)
-200
6000
6000
5000
5000
4000
4000
VFA
3000
2000
2000
BA
1000
0
3000
1000
0
50 100 150 200 250 300 350 400 450 500 550 600
Time (days)
Bicarbonate alkalinity (mgCaCO3/L)
8.0
Time (day)
Figure 3.3. Performance parameters of the UASB reactor, in different phases (according to
table 3.1). VFA: volatile fatty acids; BA: bicarbonate alkalinity.
The optimum pH for SRB growth is around 7 and lower values (pH < 5) affect bacterial
growth (Barton, 1995). Furthermore the VFA accumulation and alkalinity production, both
resulting from organic matter degradation, will define the effluent pH (figure 3.3a). Figure
3.3b shows that VFA concentration steadily increased from 2000 mg/L to 4000 mg/L which is
neutralized by alkalinity (~1500mg/L) produced during lactate oxidation, which is sufficient
for maintaining the pH in a range that enables SRB growth. The pH inside the reactor
remained fairly constant up to phase III (6.8) and a slight reduction was noticed as a result of
higher VFA production in phases IV (pH = 6.6) and V (pH = 6.4), likely derived from
63
increased organic load (figure 3.3a). The pH increased to values above 7, when recirculation
was started (phase VI), in which more alkalinity was produced as compared to the previous
phases. This behavior is a result of a higher sulfate reduction in this latter phase (table 3.3,
reaction 2); thereby recirculation can be an alternative to external alkalinity addition to
maintain pH conditions suitable for SRB development. The solution redox potential reflects
the reducing conditions in the UASB reactor (figure 3.3a) and decreased as sulfate reduction
improved. It also enabled a quick assessment of the operational conditions of the reactor.
Table 3.3. Anaerobic degradation reactions relevant to this study.
#
1
2
3
4
5
6
7
8
9
Chemical reaction
2lactate + 3SO,
+ → 6HCO0 + 3HS + H
,
2lactate + SO+ → 2acetate + 2HCO0 + HS + H
3lactate → acetate + 2propionate + HCO
0 + H
propionate + SO,
+ + H , → acetate + 2HCO0 + HS + H, O
,
acetate + SO+ → 2HCO0 + HS
4H, + SO,
+ + H → HS + 4H, O
4H, + HCO0 + H
→ CH+ + 3H, O
acetate + H, O → CH+ + HCO
0
propionate + 3H, O → acetate + HCO
0 + H + 3H,
∆G0’ (KJ)
-225.3
-160.1
-169.7
-75.8
-47.8
-151.9
-153.6
-31.0
+76.1
COD Removal efficiency and sulfate reduction
In a previous work with this inoculum, a COD/sulfate mass ratio of 2.0 was optimum for the
bacterial growth and sulfate reduction in batch conditions, resulting in 98% sulfate removal
(Barbosa et al., 2009). Therefore, this value was chosen for the phase I (enrichment), although
the actual values were slightly smaller (table 3.1). Afterwards, during phases III to V, the
organic loading was changed according to both SRB concentration and sulfate reduction
yields in the previous phases. These variations are shown in figure 3.4. During phases I
(3.48±0.33 kgCOD/m3.d) and II (4.87±0.30 kgCOD/m3.d) (figure 3.4a), the average organic
64
matter consumption was 25% and 22% respectively (figure 3.4b). Furthermore, for an SRB
population of 5.3 x 106 SRBcells/mL (figure 3.2), 51% of the oxidized organic matter was
utilized for a 36% sulfate reduction (phase I). Similarly, during phase II, a population of
8.8x107 SRBcells/mL utilized 58% of the transferred electrons for a 49% sulfate reduction. At
phase III, the SRB population increased considerably (9.5x109 SRBcells/mL) and removed
40% of the COD and 65% of the sulfate, implying that 60% of the electron flux was utilized
for sulfate reduction at a rate of 1.29 kgSO42-/m3.d (figure 3.4c). Sulfate reduction improved
up to phase IV (88%) (figure 3.4d) but the organic loading of 4.65±0.30 kgCOD/m3.d resulted
in a higher VFA production, i.e. increased lactate fermentation, which decreased sulfate
reduction to a minimum of 32% at the later stages of this phase. This behavior is consistent
with the work of Ren et al. (2007). At an even higher organic loading of 5.89 kgCOD/m3.d
(phase V), no improvement on reactor performance was observed and a reduction on COD
consumption to 23%, coupled with a sulfate reduction efficiency that varied between 39% and
72%, was noticed. Analysis of the COD consumption as a function of the organic load is
carried out in the next section.
65
(A)
I
II
III
IV
V
OLRIN
6.0
I
VI
4.8
3.6
2.4
COD removal (%)
Influent and effluent - OLR
3
(KgCOD/m .d)
7.2
(B)
100
OLR OUT
1.2
90
II
III
IV
V
VI
80
70
60
50
40
30
20
10
0.0
0
(D )
10 0
2-
3
2.0
1.5
SLR O U T
1.0
0.5
Sulfate reduction (%)
90
SLR IN
2.5
(KgSO4 /m .d)
Influent and effluent-sulfate loading rates
(C)
3.0
80
70
60
50
40
30
20
10
0
0.0
50 100 150 200 250 300 350 400 450 500 550 600
50
10 0 1 5 0 20 0 2 5 0 3 0 0 3 50 4 0 0 4 50 5 00 5 5 0 6 00
T im e (day)
Figure 3.4. Parameters monitored during sulfate reduction with lactate. Volumetric organic loading (A); COD removal efficiency (B).
Volumetric sulfate loading (C). Sulfate removal efficiency (D), during the phases I to VI (according to table 3.1).
66
As stated, up to phase V, increased lactate concentration did not improve sulfate reduction,
as the residual sulfate concentration of the reactor effluent was fairly high (780 mg/L, on
the average). Alternatively, it was decided to recirculate the effluent and this is represented
by phase VI. For 5.0 kgCOD/m3.d (similar to that applied in phase IV), reactor
performance improvement was clear as sulfate reduction increased from 65% to 89% (235
mg/L residual sulfate) for a COD consumption of 41%. This value is consistent with the
work of Kaksonen et al. (2003a) under similar conditions.
Volatile fatty acids profile and sulfate reduction
Microbial species identified in the inoculum and on the reactor biomass are depicted in
table 3.2. Furthermore, table 3.4 presents data of fed lactate (determined from the
measured COD) and measured sulfate, acetate and propionate concentrations during the
different phases. Table 3.4 also shows the fraction of lactate utilized during sulfate
reduction as predicted by reaction 2, table 3.3. The selection of reaction 2 as the main
sulfate reduction pathway was supported by alkalinity measured (figure 3.3b) during the
experiment, which would be two times higher if direct oxidation to bicarbonate was the
main reaction (reaction 1, table 3.3). From the five VFAs analyzed, only acetate and
propionate were detected and accounted for the high VFA concentrations observed in the
UASB reactor (figure 3.3b). This result is consistent with the works of Zhao et al. (2008)
and Lopes et al. (2010), which also observed sulfate reduction under acidogenic
conditions. The implications of these parameters on the reactor performance are discussed
in the following paragraphs.
67
Table 3.4.Parameters related to sulfate removal in the UASB reactor as a function of feed
lactate concentration.
Parameters
Phases monitored
Unit
Influent lactate (*)
mmol/L
I
37.24
II
52.57
III
38.06
IV
49.81
V
63.43
VI
54.16
Sulfate removal
mmol/L
7.25
9.95
12.89
13.96
12.13
17.23
Acetate analytical
mmol/L
18.41
43.33
25.07
28.83
38.66
47.00
Propionate analytical
mmol/L
12.87
19.83
9.06
19.52
37.88
7.74
Oxidized lactate (**)
%
40
38
68
56
38
64
* Determined from the measured COD in the reactor feed.
** Lactate oxidized was calculated from reduced sulfate according to reaction 2 (Table
3.3) divided by the influent lactate concentration.
The parameters shown in table 3.4 and figure 3.2 along with the microbial characterization
(table 3.2) indicate two metabolic pathways for lactate degradation (figure 3.5): (i) lactate
is first oxidized to pyruvate following to acetate by incomplete-oxidizers SRB; in which
Desulfovibrio sp. plays a key role (reaction 2, table 3.3); (ii) lactate is fermented by the
propionate CoA-transferase enzyme produced by fermenting bacteria such as Clostridium
sp(reaction 3, table 3.3) (Barton, 1995; García, 1982). These observations are supported by
the acetate and propionate accumulation in the reactor (table 3.4). Moreover, the results
suggest that the predominant metabolic pathway is defined by: (i) lactate (or COD)/sulfate
mol ratio, (ii) SRB population (bacterial counts) and (iii) the reactor hydrodynamics.
Figure 3.6a depicts both the measured acetate concentration and that predicted by reaction
2 (determined from the reduced sulfate), table 3.4. This latter was added to that produce by
reaction 3, assuming that the lactate not consumed by reaction 2 was converted to acetate.
Unlike phase II, the values predicted by the two metabolic pathways are in agreement with
the measured concentrations. Similarly, figure 3.6b presents actual and predicted (reaction
3, table 3.3) propionate concentrations. Fairly good agreement between the experimental
68
results and the predicted concentrations during phases I to IV was also noticed. The higher
than expected propionate concentration observed in phases IV and V suggests that reaction
2 itself does not account for sulfate reduction. It is likely that hydrogen produced during
lactate fermentation mainly in phase V (Garcia et al., 2001) was used as an electron donor
for sulfate reduction (reaction 6). In the study of Hwang et al. (2009) hydrogen gas
production was not significant at lower COD/sulfate ratios, but at the highest organic
loading, it became important. It must be stressed that Desulfovibrio species can grow with
both lactate and hydrogen gas as electron donors (Barton, 1995).
Figure 3.5. Metabolic pathways relevant in this study involving the anaerobic metabolism
of lactate. Species and reaction refer to identified microorganism (table 3.2) and anaerobic
degradation reactions (table 3.3).
69
Acetate concentration
(mmol/L)
OLR
5.04
OLR
4.87
55
50
45
40
OLR
3.55
35
30
25
OLR
4.65
(A)
OLR
5.89
45
(B)
40
OLR
5.89
Propionate concentration
(mmol/L)
60
OLR
3.48
20
15
10
5
35
OLR
4.87
30
25
20
OLR
4.65
OLR
3.48
OLR
5.04
OLR
3.55
15
10
5
0
I
II
III
IV
V
VI
Phases
Estimated acetate(Sulfidogenesis and fermantation products)
Analytical acetate concentration
0
I
II
III
IV
V
VI
Phases
Estimated propionate (fermantation product)
Analytical propionate conc.
Figure 3.6. Comparison between estimated and analytical concentrations of: (A) acetate;
(B) propionate, during phases I to VI (table 3.1). OLR: organic load rate (kg/m3.d).
When reactor performances with the same lactate concentration in the feed are compared
(phases I and III – 37 mmol/L lactate), it was noticed that the predominant metabolic
pathway during lactate degradation (oxidation or fermentation) was defined by the SRB
population. In phase I, a low population of 5.3x106 SRBcells/mL (Fig. 3.2) used 40% of
the fed lactate (37 mmol/L) to reduce 7.25 mmol/L of the sulfate (following reaction 2,
table 3.3). The remaining lactate (22.4 mmol/L) was thus fermented and produced
propionate (reaction 3, table 3.3), which is consistent with the measured propionate
concentration (12.9±3.3 mmol/L) as shown in figure 3.6(b). Therefore, during phase I, the
dominant reaction in this reactor seemed to be lactate fermentation to propionate which
may have been carried out by Clostridium sp (table 3.2). In this phase, the measured
acetate concentrations (18.41±3.2 mmol/L, fig. 3.6(a)) are similar to that resulting from
reactions 2 and 3 occurring in the reactor. The incomplete lactate oxidation by SRB
(reaction 2) would produce 14.5 mmol/L acetate and lactate fermentation (reactor 3)
another 7.4 mmol/L acetate.
70
Likewise during phase III and at the same lactate concentration, but with a higher SRB
population (9.5x109 SRBcells/mL), reaction 2 predicts that reducing 12.89 mmol/L of
sulfate (analytical value), would require 25.8 mmol/L (68% of the initial concentration) of
lactate and 25.8 mmol/L of acetate would also be produced. The measured acetate
concentration (25.2±6.8 mmol/L) is consistent with such analysis. Analyzing reaction 3 it
can be predicted that the remaining lactate (12.3 mmol/L) was fermented to produce 4.1
mmol/L of acetate (reaction 3) and 8.2 mmol/L propionate and this pathway is also
supported by the measured propionate concentration (9.1 mmol/L).
Organic loading effects on sulfate reduction and microbial population can be assessed by
comparing phases III to V, since both parameters are known to affect sulfate reduction
(Lens et al., 2003; Reis et al., 1988; Sipma et al., 1999). In these three phases, the SRB
population stabilized within the 10 8 - 10 9 cells/mL range and it could be inferred that an
increase in lactate concentration did not result in larger sulfate reduction yields, since the
sulfate reduction was higher in phase IV (4.8±0.4 g/L COD) than in phase V, when the
COD was increased to 6.0 g/L (table 3.1 and figure 3.4). During phase III, 68% of the fed
lactate (38 mmol/L) was used to reduce 13 mmol/L of the sulfate. However, as the lactate
concentration was increased to 50 mmol/L (phase IV), the sulfate reduction represented
56% of lactate degradation; whereas at phase V, when lactate concentration increased
further to 63 mmol/L, only 38% the lactate oxidation seemed to be coupled to the reduction
of 12 mmol/L of the sulfate by SRB. Such results show that even at high organic loadings,
sulfate reduction is not complete and it seems that fermentation is being promoted; i.e. as
the COD/Sulfate mass ratio increased, the sulfate reduction rate also increased, but the
fraction of the organic matter effectively used for sulfate reduction was reduced, indicating
that the organic matter was being fermented (SRB or Clostridium), which had also been
71
observed by Ren et al. (2007). MPM, which were proposed by many works (Kalyuzhnyi
and Fedorovich, 1998; O'Flaherty et al., 1998; Omil et al., 1998) as the main competitors
with SRB for substrate oxidation, do not seem to be important in the present work. The
third effect accounting for the predominant metabolic pathway is recirculation that will be
discussed in the next part of the present work.
Table 3.2 indicates the absence of SRB that oxidizes lactate to CO2(Barton, 1995) in the
reactor so that reaction 1 (table 3.3) was not expected to occur. Analyzing the SRB
population, only one acetate-oxidizer SRB was detected (Desulfobacter halotolerans).
Furthermore, methanogens were also observed and both (methanogens and D.
halotolerans) can oxidize acetate. Nevertheless, as acetate accumulated in the effluent
during the sulfate reduction in all of the phases (I-VI), it can be inferred that none of them
predominated in the reactor. This would explain the high acetate concentrations observed
(1100-2600 mg/L) in the effluent and is consistent with the work of Lopes et al. (2010),
who studied thermophilic sulfate reduction with sucrose (2.0 g/L) as electron donor. As the
kinetics of sulfate reduction and organic matter oxidation is compared, acetate oxidation is
the limiting step during both sulfate reduction and methane production (Aquino and
Chernicaro, 2005; Colleran et al., 1995; Kaksonen et al., 2006). This behavior is because
both acetate-oxidizer SRB and acetate-reducing MPM have small growth rates i.e., specific
growth rate (µ max) values that vary from 0.002 to 0.068 h-1 for the former (Elferink et al.,
1998; Lawrence and Marchant, 1991) and from 0.0046 to 0.01 h-1 for MPM (Kalyuzhnyi
and Fedorovich, 1998; Zuhair et al., 2008). For sulfate reduction via incomplete lactate
oxidation, much higher µ max values are reported, ranging from 0.23 to 0.498 h-1(Kaksonen
et al., 2003b; Widdel, 1988; Zuhair et al., 2008). Therefore kinetic effects account for the
acetate presence in the reactor for the same residence time. As figure 3.6a shows, the
72
acetate concentrations predicted by reactions 2 and 3 (table 3.3), are similar to that actually
measured, supporting the proposed metabolic pathways.
MPM growth is also partially inhibited by high sulfide concentrations during
sulfidogenesis(Dar et al., 2008; Kaksonen et al., 2003a; Nagpal et al., 2000; Omil et al.,
1998). However, different species have different tolerances to total sulfide and Chen
(2008) suggests that sulfide inhibition ranges from 100 mg/L to 800 mg/L in the case of the
bissulfide ion (HS-) or 50-400 mg/L when the predominant species is H2S. He also
indicated that fermentative bacteria are less affected by the sulfide concentration than
acetoclastic SRB and methanogens. Considering the SRB group, incomplete oxidizers SRB
are less affected by sulfide than complete oxidizers SRB (Kaksonen and Puhakka, 2007).
In the present study, the pH inside the UASB reactor remained in the 6.4-6.8 range;
thereby some 25% and 75% of the total sulfide was estimated to be present as HS- and
H2S, respectively. Moreover, during phases IV and V, the total sulfide concentration varied
from 184 mg/L to 250 mg/L, which was smaller than the value predicted when the sulfate
reduction yields were analyzed. This lower than expected value could be a result of H2S
volatilization during the experimental runs (parameter not followed). Nevertheless, the
measured H2S concentration is high enough to inhibit acetate oxidation by either
acetoclastic SRB or methanogens. For instance it has been shown that 270 mgH2S/L and
160 mgH2S/L accounted for a 50% inhibition on acetoclastic SRB and methanogens
growth, respectively (Yamaguchi et al., 1999).
Propionate presence in the reactor effluent is an indication of fermenting microorganisms
(Dar et al., 2008), which also produces acetate. Assuming the predominance of incomplete
oxidizers SRB in the reactor, an estimate of propionate concentration can be performed
73
from the lactate concentration not used for sulfate reduction (table 3.4), since lactate was
not detected in the reactor effluent. The absence of lactate can be verified by equation3.1
that presents a mass balance for the COD measured in the system.
CODremoved = CODin – CODout
(3.1)
and
CODout = CODVFA + CODTDS + CODresidual
(3.2)
Where CODin and CODout are the chemical oxygen demand in the reactor influent and
effluent, respectively; CODVFA corresponds to VFA (acetate and propionate solely);
CODTDS, to sulfide ions and CODresidual, to any lactate remaining in the system. Sulfide
ions were removed from the reactor effluent sample by acidification and stripping (CelisGarcía et al., 2007); whereas bacterial cells did not contribute to the measured COD, due to
membrane filtration prior to analysis. Therefore, CODout is achieved from CODVFA and
CODresidual, as shown in equation 3.2. It can be noticed in figure 3.7, that the CODout value
is similar or smaller than CODVFA; whereupon no residual lactate was expected. It was
likely degraded to either acetate (by either SRB or fermentative microorganisms, or both)
or propionate (fermentative microorganisms).
74
7000
Chemical oxygen demand
(mg/L)
CODout
CODVFA
6000
5000
4000
3000
2000
1000
0
I
II
III
IV
V
VI
Phases
Figure 3.7. Values of effluent- and CODVFA during sulfate reduction in the UASB reactor.
CODout was measured and CODVFA values were determined from the measured propionate
and acetate concentrations.
By analyzing the propionate profile (figure 3.6b), its concentration in the phases where
SRB population was low (I and II) could be predicted (assuming that the lactate not
consumed by reaction 2 was converted to propionate by reaction 3) fairly accurately, i.e.
lactate fermentation to acetate and propionate. Conversely, in the phases where the SRB
population was high (>108 cells/mL), propionate concentration increased (from 9.06
mmol/L in phase III to 37.88 mmol/L in phase V) with the organic load, suggesting a high
fermentative activity that decreased the specific sulfate reduction rate from values above
0.075 gSO42-/gSSV.d to lower yields (0.060 gSO42-/gSSV.d). Similar behavior was
observed by Dar et al. (2008) who detected propionate at a low sulfate concentration (20
mM lactate and 10.3 mM sulfate), indicating that lactate fermented. Oyekola (2009) also
observed similar results when the reactor dilution rate was high, which was related to faster
lactate fermentation as compared to sulfate reduction.
75
SRB can degrade organic compounds by diverse metabolic pathways but most
Desulfovibrio species do not grow in the presence of volatile fatty acids (Barton, 1995).
Therefore, increased propionate concentration indicates lower growth rates of
Desulfobulbus, the only propionate-degrading SRB identified in the UASB reactor.
Another possibility would be propionate degradation by syntrophic microorganisms
following reaction 9 (table 3.3). However, this reaction is not supposed to occur in the
presence of high acetate concentrations (Aquino and Chernicaro, 2005). Notwithstanding,
propionate concentration was decreased when recirculation was applied as will be
discussed next.
3.3.3. Influence of effluent recirculation in sulfate reduction
As shown in figure 3.6, when the organic loading was high (phases III, IV and V), there
was propionate build-up (due to lactate fermentation), which may be ascribed to poor
sludge granulation and a low upflow velocity in the UASB reactor (Steed et al., 2000).
With recirculation (Phase VI), propionate concentration reduced from almost 40mmol/L, at
phase V (5.89 kg/m3.d) to values below 10mmol/L at a similar organic loading rate (5.04
kg/m3.d). This latter propionate concentration is similar to that observed during phase III
(~10 mmol/L), when the organic loading was only 3.55 kg/m3.d. As already stated,
Desulfobulbus can degrade propionate (reaction 4, table 3.3), but its growth rate is small in the range 0.037-0.11 h-1(Colleran et al., 1998; Kalyuzhnyi and Fedorovich, 1998) - as
compared to those bacteria belonging to the Desulfovibrio genera. The applied
recirculation rate was 93, which resulted in increased upflow velocities (from 0.02 m/h to
1.75 m/h), thereby improving the mass transfer in the reactor. Considering reaction 2 (table
3.3) as the main metabolic pathway for lactate oxidation during sulfate reduction (89%),
76
along with reaction 3, and assuming that the residual lactate had been used for the
propionate production, nearly 1017 mg/L propionate was expected in the reactor. As the
measured propionate concentration was 596 ± 188 mg/L, it is therefore proposed that the
propionate oxidation by Desulfobulbus became relevant in this phase. Lactate fermentation
to propionate followed by its oxidation by Desulfobulbus was also observed by Dar et al.
(2008) and Zhao et al. (2008) who attributed high sulfate reduction yields to propionate
oxidation.
The present work suggests that regardless of the substrate being degraded, the performance
of an UASB reactor can be followed by the VFA profile. For instance, the propionate
presence could be taken as an indication of fermentation, which is an important metabolic
pathway and this would imply in low sulfate reduction efficiency. Ideally, an UASB
reactor for sulfate reduction would not produce propionate at optimum sulfate reduction
conditions. Nevertheless, the residual COD was high and requires downstream treatment
but it is likely to be easily degradable, due to the presence of only acetate (47 mmol/L) and
propionate (8 mmol/L) when lactate was the only carbon source.
3.4. Conclusions
This work showed that the best results of a UASB reactor for sulfate reduction were
achieved for COD/sulfate mass ratios in the 1.7-2.5 range, in the presence of fermentative,
methanogenic microorganism and for a low diversity of acetoclastic bacteria, but enriched
in an incompletely-oxidizing SRB. At either low SRB concentration (106 cells/mL) or high
organic loading (6.0 kgCOD/m3.d), fermentation was the main metabolic pathway;
thereby, acetate and propionate were predominant in the reactor. Conversely, for a high
77
SRB population (109 cells/mL) and lower organic loading (3.55COD kg/m3.d.), sulfate
reduction becomes the main metabolic pathway and an average sulfate reduction value of
66% was observed without recirculation in the reactor treating 2.0 g/L sulfate for a 24
hours residence time. Sulfate reduction increased to 89% (0.087 gSO42-/gSSV.d) and
resulted in an increase in acetate concentrations couple to low propionate content as
recirculation was applied. High propionate levels are an indication of lactate fermentation
whereupon sulfate reduction was impaired.
The financial support from the funding agencies FINEP, FAPEMIG, CNPq, CAPES as
well as Vale is gratefully appreciated. The “Conselho Nacional de Pesquisas - CNPq
scholarships to S. M. Bertolino and S. F. Aquino R. Guerra-Sá, V. A. Leão are especially
acknowledged.
3.6. References
APHA, 2005.Standard Methods for the examination of water and wastewater, Washington,
DC.
Aquino, S.F., Chernicaro, C.A., 2005. Build up of volatile fatty acids (VFA) in anaerobic
reactors under stress conditions: causes and control strategies. Engenharia Sanitaria e
Ambiental 10.
Barbosa, L.P., Bertolino, S.M., Freitas, P.C., Oliveira, V.A., Pina, P.D., Leão, V.A.,
Teixeira, M.C., 2009. Effects of differnt COD/sulfate ratios on the growth of metal tolerant
sulfate reducing bacteria (SRB), biohydrometallurgy: a meeting point between microbial
78
ecology, metal recovery process and environmental remediation., Proceedings of
Biohydrometallurgy, Bariloche, pp. 569-572.
Barton, L.L., 1995. Sulfate-reducing bacteria. Plenum Press, New York.
Beaulieu, S., Zagury, G.J., Samson, R., 2000. Bioactivation and bioaugmentation of a
passive reactor for acid mine drainage treatment, in: Singhal, R.K., Mehrotra, A.K. (Eds.),
Environmental Issues and Management of Waste in Energy and Mineral Production,
Rotterdam, The Netherlands, pp. 533-537.
Bhattacharya, S.K., Uberoi, V., Dronamraju, M.M., 1996. Interaction between acetate fed
sulfate reducers and methanogens. Water Research 30, 2239-2246.
Briones, A.M., Daugherty, B.J., Angenent, L.T., Rausch, K., Tumbleson, M., Raskin, L.,
2009. Characterization of microbial trophic structures of two anaerobic bioreactors
processing sulfate-rich waste streams. Water Research 43, 4451-4460.
Celis-García, L.B., Razo-Flores, E., Monroy, O., 2007. Performance of a down-flow
fluidized bed reactor under sulfate reduction conditions using volatile fatty acids as
electron donors. Biotechnology and Bioengineering 97, 771-779.
Chen, Y., Cheng, J.J., Creamer, K.S., 2008. Inhibition of anaerobic digestion process: A
review. Bioresource Technology 99, 4044-4064.
Colleran, E., Finnegan, S., Lens, P., 1995. Anaerobic treatment of sulphate-containing
waste streams. Antonie van Leeuwenhoek 67, 29-46.
Colleran, E., Pender, S., Philpott, U., O'Flaherty, V., Leahy, B., 1998. Full-scale and
laboratory-scale anaerobic treatment of citric acid production wastewater. Biodegradation
9, 233-245.
Dar, S., Kleerebezem, R., Stams, A., Kuenen, J., Muyzer, G., 2008.Competition and
coexistence of sulfate-reducing bacteria, acetogens and methanogens in a lab-scale
anaerobic bioreactor as affected by changing substrate to sulfate ratio. Applied
Microbiology and Biotechnology 78, 1045-1055.
79
Elferink, S.J.W.H.O., Luppens, S.B.I., Marcelis, C.L.M., Stams, A.J.M., 1998. Kinetics of
Acetate Oxidation by Two Sulfate Reducers Isolated from Anaerobic Granular Sludge.
Appl. Environ. Microbiol. 64, 2301-2303.
Garcia, C., Moreno, B.A., Ballester, A., Blázques, M.L., González, F., 2001.
Bioremediation of an industrial acid mine water by metal tolerant suphate-reducing
bacteria. Minerals Engineering 14, 997-1008.
García, J.L., 1982. Relations between acidogenesis and the utilization of lactate, sulfate
and nitrate during anaerobic digestion, Simposio Panamericano de Combustibles y
Productos Químicos Vía Fermentación "Avances en Digestión Anaeróbica", Mexico City,
Mexico, pp. 20–30.
Gonçalves, M.M.M., da Costa, A.C.A., Leite, S.G.F., Sant’Anna Jr, G.L., 2007. Heavy
metal removal from synthetic wastewaters in an anaerobic bioreactor using stillage from
ethanol distilleries as a carbon source. Chemosphere 69, 1815-1820.
Harada, H., Uemura, S., Momonoi, K., 1994. Interaction between sulfate-reducing bacteria
and methane-producing bacteria in UASB reactors fed with low strength wastes containing
different levels of sulfate. Water Research 28, 355-367.
Hwang, J.-H., Cha, G.-C., Jeong, T.-Y., Kim, D.-J., Bhatnagar, A., Min, B., Song, H.,
Choi, J.-A., Lee, J.-H., Jeong, D.-W., Chung, H.-K., Park, Y.-T., Choi, J., Abou-Shanab,
R.A.I., Eun Oh, S., Jeon, B.-H., 2009. Effect of COD/SO42- ratio and Fe(II) under the
variable hydraulic retention time (HRT) on fermentative hydrogen production. Water
Research 43, 3525-3533.
INAP, 2003.Treatment of sulphate in mine effluentes.International network for acid
prevention, p. 129.
Kaksonen, A., Franzmann, P., Puhakka, J., 2003a. Performance and Ethanol Oxidation
Kinetics of a Sulfate-Reducing Fluidized-Bed Reactor Treating Acidic Metal-Containing
Wastewater. Biodegradation 14, 207-217.
80
Kaksonen, A.H., Plumb, J.J., Franzmann, P.D., Puhakka, J.A., 2004. Simple organic
electron donors support diverse sulfate-reducing communities in fluidized-bed reactors
treating acidic metal- and sulfate-containing wastewater. FEMS Microbiology Ecology 47,
279-289.
Kaksonen, A.H., Plumb, J.J., Robertson, W.J., Riekkola-Vanhanen, M., Franzmann, P.D.,
Puhakka, J.A., 2006. The performance, kinetics and microbiology of sulfidogenic
fluidized-bed
treatment
of
acidic
metal-
and
sulfate-containing
wastewater.
Hydrometallurgy 83, 204-213.
Kaksonen, A.H., Puhakka, J.A., 2007. Sulfate Reduction Based Bioprocesses for the
Treatment of Acid Mine Drainage and the Recovery of Metals. Engineering in Life
Sciences 7, 541-564.
Kaksonen, A.H., Riekkola-Vanhanen, M., Puhakka, J.A., 2003b.Optimization of metal
sulphide precipitation in fluidized-bed treatment of acidic wastewater. Hydrometallurgy
37, 255-266.
Kalyuzhnyi, S.V., Fedorovich, V.V., 1998. Mathematical modelling of competition
between sulphate reduction and methanogenesis in anaerobic reactors. Bioresource
Technology 65, 227-242.
Kosinska, K., Miskiewicz, T., 2009.Performance of an anaerobic bioreactor with biomass
recycling, continuously removing COD and sulphate from industrial wastes. Bioresource
Technology 100, 86-90.
Lawrence, R.W., Marchant, P.M., 1991. Acid Rock Drainage Prediction Manual: A
Manual of Chemical Evaluation Procedures for the Prediction of Acid Generation from
Mine Wastes. CANMET (Canadian Centre for Mineral and Energy Technology),
Vancouver, BC, Canada, p. 47.
Lens, P.N.L., Gastesi, R., Lettinga, G., 2003. Use of Sulfate Reducing Cell Suspension
Bioreactors for the Treatment of SO2 Rich Flue Gases. Biodegradation 14, 229-240.
81
Lettinga, G., Hulshoff Pol, L.W., 1991. UASB Process design for various types of
wastewaters, Proc. IAWPRC Int. Specialized Workshop, Valladolid, 1990. Water Sci.
Technol. 24,8 (1991) 87-109.
Liamleam, W., Annachhatre, A.P., 2007.Electron donors for biological sulfate reduction.
Biotechnology Advances 25, 452-463.
Lopes, S.I.C., Capela, M.I., Lens, P.N.L., 2010.Sulfate reduction during the acidification of
sucrose at pH 5 under thermophilic (55 °C) conditions. II: Effect of sulfide and COD/ SO 4
2 - ratio. Bioresource Technology 101, 4278-4284.
Lopes, S.I.C., Wang, X., Capela, M.I., Lens, P.N.L., 2007. Effect of COD/SO42- ratio and
sulfide on thermophilic (55°C) sulfate reduction during the acidification of sucrose at pH
6.Water Research 41, 2379-2392.
Mizuno, O., Li, Y.Y., Noike, T., 1998.The behavior of sulfate-reducing bacteria in
acidogenic phase of anaerobic digestion. Water Research 32, 1626-1634.
Mohan, S.V., Prasad, K.K., Rao, N.C., Bhaskar, Y.v., Babu, V.L., Rajagopal, D. and
Sarma, P.N., 2005.Biological treatment of low-biodegradable composite chemical
wastewater using upflow anaerobic sludge blanket (UASB) reactor.Journal of scientific
and industrial research 64, 771-777.
Nagpal, S., Chuichulcherm, S., Livingston, A., Peeva, L., 2000. Ethanol utilization by
sulfate-reducing bacteria: An experimental and modeling study. Biotechnology and
Bioengineering 70, 533-543.
O'Flaherty, V., Lens, P., Leahy, B., Colleran, E., 1998.Long-term competition between
sulphate-reducing and methane-producing bacteria during full-scale anaerobic treatment of
citric acid production wastewater. Water Research 32, 815-825.
Omil, F., Lens, P., Visser, A., Hulshoff Pol, L.W., Lettinga, G., 1998. Long-term
competition between sulfate reducing and methanogenic bacteria in UASB reactors
treating volatile fatty acids. Biotechnology and Bioengineering 57, 676-685.
82
Oyekola, O.O., van Hille, R.P., Harrison, S.T.L., 2009. Study of anaerobic lactate
metabolism under biosulfidogenic conditions.Water Research 43, 3345-3354.
Poinapen, J., Ekama, G., Wentzel, M., 2009a. Biological sulphate reduction with primary
sewage sludge in an upflow anaerobic sludge bed (UASB) reactor - Part 2: Modification of
simple wet chemistry analytical procedures to achieve COD and S mass balances. Water
SA 35, 535-542.
Poinapen, J., Wentzel, M.C., Ekama, G.A., 2009b. Biological sulphate reduction with
primary sewage sludge in an upflow anaerobic sludge bed (UASB) reactor - Part 1:
Feasibility study. Water SA 35 525-534.
Postgate, J.R., 1963.Versatile medium for the enumeration of sulfate-reducing
bacteria.Applied Microbiology 11, 265-267.
Rampinelli, L., Azevedo, R., Teixeira, M., Guerra-Sá, R., Leão, V., 2008. A sulfatereducing bacterium with unusual growing capacity in moderately acidic conditions.
Biodegradation 19, 613-619.
Reis, M.A.M., Gonçalves, L.M.D., Carrondo, M.J.T., 1988. Sulfate Reduction in
Acidogenic Phase Anaerobic Digestion. Water Science & Technology 20, 345–351.
Ren, N.-Q., Chua, H., Chan, S.-Y., Tsang, Y.-F., Sin, N., 2007.Effects of COD/SO42Ratios on an Acidogenic Sulfate-Reducing Reactor. Industrial & Engineering Chemistry
Research 46, 1661-1666.
Shayegan, J., Ghavipanjeh, F., Mirjafari, P., 2005.The effect of influent COD and upward
flow velocity on the behaviour of sulphate-reducing bacteria.Process Biochemistry 40,
2305-2310.
Sipma, J., Lens, P., Vieira, A., Miron, Y., van Lier, J.B., Hulshoff Pol, L.W., Lettinga, G.,
1999. Thermophilic sulphate reduction in upflow anaerobic sludge bed reactors under
acidifying conditions. Process Biochemistry 35, 509-522.
83
Steed, V.S., Suidan, M.T., Gupta, M., Miyahara, T., Acheson, C.M., Sayles, G.D.,
2000.Development of a sulfate-reducing biological process to remove heavy metals from
acid mine drainage. Water Environment Research 72, 530-535.
USEPA, 2009.Edition of the Drinking Water Standards and Health Advisiories Tables.U.S.
Environmetal Protection Agency, Waschington.DC., p. 18.
Vallero, M.V.G., Treviño, R.H.M., Paulo, P.L., Lettinga, G., Lens, P.N.L., 2003. Effect of
sulfate on methanol degradation in thermophilic (55 °C) methanogenic UASB reactors.
Enzyme and Microbial Technology 32, 676-687.
WHO, 2011. Guidelines for drinking-water quality, 4rd ed, Genebra, p. 564.
Widdel, F., 1988.Microbiology and ecology of Sulfate- and sulfur- reducing bacteria
Biology of anaerobic microorganisms. John Willey & Sons, New York, pp. 469-585.
Yamaguchi, T., Harada, H., Hisano, T., Yamazaki, S., Tseng, I.C., 1999.Process behavior
of UASB reactor treating a wastewater containing high strength sulfate.Water Research 33,
3182-3190.
Zhao, Y., Ren, N., Wang, A., 2008. Contributions of fermentative acidogenic bacteria and
sulfate-reducing bacteria to lactate degradation and sulfate reduction. Chemosphere 72,
233-242.
Zuhair, S.A., El-Naas, M.H., Hassani, H.A., 2008. Sulfate inhibition effect on sulfate
reducing bacteria. Journal of Biochemical Technology 1, 39-44.
84
CAPÍTULO 4
GLYCEROL AS ANELECTRON DONOR FOR SULFATE REDUCTION IN
FLUIDIZED BED REACTORS
Abstract
One of the greatest challenges to the full implementation of biological sulfate reduction is
the cost of the electron source, along with its availability. Crude glycerol, a by-product of
biodiesel production, is one such source. The performance of pure glycerol (the main
constituent of crude glycerol) as a carbon and electron source for sulfate reduction was
compared to that achieved with lactate. Continuous experiments were performed in a
fluidized bed reactor (FBR) containing activated carbon as a carrier for a mixed bacterial
population composed of sulfate-reducing and fermentative bacteria. Incomplete lactate
oxidation was the main metabolic pathway during sulfate reduction (70-90%), which
resulted in large acetate concentrations in the FBR effluent (2500mg COD/L). Conversely,
pure glycerol degradation by a syntrophic cooperation between Desulfovibrio spp. and
Clostridium spp. produced a residual sulfate concentration of 254mg/L and an acetate
concentration that was 2.5 times lower than that obtained with lactate. Since glycerol was
proved to oxidized by Sulfate Reducing Bacteria (SRB), crude glycerol may be foreseen as
a cost-effective alternative to lactate and for sulfate reduction and a marketable solution for
the glycerol residual generated by the biodiesel industry.
Key-words: glycerol, sulfate-reducing bacteria, Clostridium, Desulfovibrio, fluidized bed
reactor.
85
4.1. Introduction
Despite all efforts, recent figures have shown an increase in carbon dioxide emissions in
the year 2010. Such an outcome was ascribed to an increase in fossil fuel consumption,
which is considered the main cause of global warming. Biofuels are among the renewable
alternatives to fossil fuels, and biodiesel production in particular has received much
attention in recent years.
The widespread embrace of biodiesel requires addressing relevant environmental issues
related to its production, such as the production of impure glycerol (g-phase), an important
by-product of the process. The composition of the g-phase varies widely depending on the
raw material, but despite its utilization in the cosmetic industry, a surplus of impure
glycerol is forecasted because of the growth of the biodiesel industry; the latter also
accounts for decreases in its price [1]. G-phase has been suggested as a carbon and energy
source for microbial growth during anaerobic digestion for methane production [2]. In this
work, it is proposed as a carbon and electron source for sulfate reduction, an alternative
which has received much less attention [3-4].
The biotechnologies for sulfate reduction have been developed throughout the past ten
years; however, only a few industrial operations have actually been commissioned [5].
Although technically feasible, industrial applications have been limited by the costs
associated with the purchase of the carbon source [5]. Table 1.2 summarizes different
studies of sulfate reduction in which different substrates were studied as carbon and
electron sources. Lactate and ethanol are the most studied organic substrates. Because of
its utilization by many sulfate-reducing bacteria (SRB), lactate is preferred for sulfate
86
reduction, although it is a more expensive carbon source than ethanol (largely applied
industrially). Crude glycerol (g-phase) would also be an inexpensive alternative to lactate
due to both its availability as a by-product of the biodiesel process and high energy
content.
Qatibi et al. [3] investigated glycerol utilization for sulfate reduction, focusing on the
metabolic pathways associated with glycerol oxidation. A mixed culture degraded glycerol
to 1,3-propanediol (1,3-PD) prior to propionate and acetate accumulation in a batch
reactor. Despite a low sulfate reduction (21%), full glycerol degradation was accompanied
by propionate production. Afterwards, Qatibi et al. [3] compared glycerol oxidation by
both Desulfovibrio carbinolicus and Desulfovibrio fructosovorans and observed that D.
carbinolicus oxidized glycerol to 3-hydroxypropionate, whereas D. fructosovorans
produced acetate. More recently, Dinkel et al. [4] studied the growth kinetics of an SRB
consortium in the presence of glycerol. The specific growth rate was determined to be
0.56d-1 for acetateproduced/sulfateremoved molar ratio of 0.4. With these outcomes, along with
the microbial diversity identified (Desulfivibrio baarsii, Desulfomicrobium sp. and
Desulfatomaculum sp.), Dinkel et al. [4] ruled out any methanogenic activity in the reactor,
whereas a syntrophic association between SRB and acidogenic microorganisms during
sulfate reduction with glycerol was proposed.
The full implementation of an anaerobic sulfate reduction process requires the
investigation of continuous operations in anaerobic reactors. FBR have the advantage of
improved mass transfer and high biomass concentrations, which enable high sulfate
reduction yields. Therefore, the main objective of this work was to develop a sulfidogenic
process using glycerol as the only carbon source for SRB growth, using lactate as a
87
positive control. Such an approach has not been described to date. The metabolic pathways
related to the microorganisms present in the bioreactor, along with the carbon sources
utilized (lactate and glycerol), were compared.
4.2. Experimental
The strategy applied in this work was to stimulate sulfate reduction with a carbon source
that can be utilized by different microbial strains, i.e., lactate. After establishing a high
biomass concentration containing predominantly SRB, the carbon source was changed to
glycerol, and the operational parameters were then followed.
4.2.1. Anaerobic reactor
Figure 4.1 shows a schematic diagram of the lab-scale FBR reactor. The total volume was
1.28L, and the FBR was placed inside a fume hood in a temperature-controlled room,
whereby the temperature was maintained at 25±2ºC. Three sampling ports, a gas outlet, a
feed tank, as well as an effluent tank completed the system. A peristaltic pump provided
solution pumping to the reactor. Activated carbon was used as the biomass carrier material
(150g; 2.1mm mean diameter; 1.63g/cm3 density, 566m2/g surface area). The carrier
material was fluidized by means of flow recirculation by a second pump at a flow rate set
at 166L/h. This resulted in a superficial velocity of 75m/h and 86% bed expansion. For
fluidization, the effluent of the top port was recycled in the system.
88
Figure 4.1. Schematic diagram of the FBR reactor for sulfate reduction.
4.2.2. Microorganisms and reactor start-up
The original inoculum (granular sludge) was obtained from an upflow anaerobic sludge
blanket UASB reactor treating domestic wastewater. The Postgate C mineral medium, used
in the start-up and continuous operation of the reactor, contained 0.5g/L KH2PO4; 1.0g/L
NH4Cl; 0.06g/L MgSO4.7H2O; 0.1g/L FeSO4.7H2O; 0.25g/L yeast extract; 2.96g/L
Na2SO4; and 3.76g/L lactate or glycerol. The FBR reactor was inoculated with 100mL of
sludge containing 38g of volatile suspended solids (VSS/L). Afterward, pumping of the
growth medium amended with sulfate was initiated in a semi-batch mode, which
89
comprised 12-h pumping and 48-h rest periods such that adaptation to the reactor was
achieved. As soon as the whole reactor was filled with medium and biomass, the
continuous operation was initiated.
4.2.3. Operational methods
After inoculation, the FRB reactor was operated at low expansion (20%) for 45 days to
avoid biomass washing and to enable colonization of the carrier material (activated
carbon). Subsequently, this reactor was operated for 580 days at a hydraulic retention time
(HRT) of 10h and an 86% fluidization rate. The reactor was fed with a synthetic effluent
(modified Postage C medium, section 4.2.2) containing 2000mg/L sulfate, and lactate or
glycerol, or both, as carbon and electron sources. Both organic and sulfate loads were
varied according to the sulfidogenic performance exhibited by the reactor. The reactor
operated at a COD/sulfate ratio that varied between 2.5 in phase I and 1.8 in phase II.
Conversely, during phases III and IV, the organic load was set to 11-12kgCOD/m3.d,
whereas the sulfate load varied according to table 4.1. During phase V, glycerol
progressively replaced lactate as the carbon and electron donor. To achieve that, the
organic load was increased to 15.21kgCOD/m3.d by glycerol (41mmol/L) addition to the
growth medium already containing lactate (44mmol/L). Throughout this phase, the lactate
concentration was reduced to zero in 61 days, whereas the glycerol concentration was
maintained constant (table 4.2).
90
Table 4.1. Operational parameters during sulfate reduction in the FBR reactor.HRTa of 10
hours and 25ºC.
Phases
I
II
III
IV
Lactate
(mmol/L)
53 ± 2.9
38 ± 2.8
52 ± 2.6
52 ± 3.1
SO42(mmol/L)
21 ± 0.85
22 ± 0.85
25 ± 1.43
29 ± 4.60
OLRb
(kg/m3 .d)
12.34 ± 0.98
8.70 ± 0.63
11.82 ± 0.58
11.74 ± 0.73
SLRc
(kg/m3 .d)
4.82 ± 0.32
4.94 ± 0.20
5.61 ± 0.29
6.46 ± 0.34
COD/SO42molar ratio
2.52±0.18
1.73 ± 0.18
2.08 ± 0.17
1.79 ± 0.09
a
HRT: hydraulic retention time.
b
OLR: organic loading rate.
c
SLR: sulfate loading rate.
Table 4.2.Operational parameters during replacement of lactate by glycerol in the FBR
reactor with SLR of 3.62 ± 0.23Kg/m3.d, at HRT of 13 hours and 25ºC.
Phase
V
Glycerol
(mmol/L)
41
41
41
41
41
Lactate
(mmol/L)
44
33
22
11
0
COD
(Kg/m3)
8.50 ± 0.40
7.74 ± 0.05
6.64 ± 0.20
5.61 ± 0.15
5.22 ± 0.25
OLR
(Kg/m3 .d)
15.21 ± 0.75
13.90 ± 0.10
11.90 ± 0.34
9.98 ± 0.26
9.21 ± 0.45
Period
Days
26
8
13
14
34
The reactor effluent was analyzed twice a week for total and filtered chemical oxygen
demand (COD), sulfate, alkalinity, volatile fatty acids (VFA), volatile suspended solids
(VSS) and pH. Once a week, a sample from inside the reactor was withdrawn to measure
VSS, alkalinity and pH, whereas viable cells were determined monthly.
4.2.4. Batch experiments with glycerol
Once steady-state conditions were attained in the FBR fed with glycerol, metabolic
pathways related to biomass activity were determined in batch experiments performed with
a biomass sample withdrawn from the reactor. These batch tests were performed at 28°C in
0.5L serum bottles closed with rubber stoppers and maintained in a bench shaker (150min91
1
). Phosphate and bicarbonate ions were utilized as buffers so as to avoid acidification. The
bacterial inoculum (10% v/v) was transferred to the flasks, which were flushed with N2,
over 24h. Then, growth medium (Postgate C) containing sulfate and glycerol (as a
substitute for lactate) was added and the experiment was run for the next 7 days.
Sulfidogenic activity was monitored by sulfate and glycerol concentrations as well as by
the production of organic acids (propionate, butyrate, acetate). Total organic carbon was
also measured to investigate glycerol degradation.
The sulfate concentration was determined by ionic chromatography (Metrohm) using an
ASSUP-10 column and conductivity detection. VFA (acetate, propionate, valerate,
butyrate) and lactate were analyzed by high-performance liquid chromatography (HPLC,
Shimadzu) with an ion-exchange column Aminex HPX-87H 300mm x 7.8mm (Bio-Rad)
according to a procedure detailed elsewhere [6]. Prior to injection, samples were filtered
using 0.22µm membrane filters (Millipore, Corp.). Bicarbonate alkalinity (BA) was
assayed by titration with 0.1M sulfuric acid solution to pH 4.0, and VSS and COD were
assayed according to the Standard Methods for Water and Wastewater [7]. Before COD
determination, any sulfide present in the samples was removed by adding a drop of HCl
(35%) and flushing for 10min with N2. The solution pH (Hanna HI931400) and redox
potential (Digimed) (versus an Ag/AgCl electrode) were also recorded. The glycerol
concentration was determined spectrophotometrically following the procedure described
by Bondioli and Bella [8]. Total dissolved organic carbon was analyzed by the persulfate
method in a HiperTOC analyzer from Thermo Scientific.
92
Microorganisms were enumerated by a three-tube most probable number (MPN) procedure
using 10-fold serial dilutions in selective media. The SRB were enumerated in a specific
medium for SRB (Postgate C) [9]. Prior to the experiments, culture tubes were degassed
with pure N2, sealed and autoclaved (120°C, 1.5atm, 20min). Then, the culture tubes and
control tube were incubated for 30 days at 35°C. Cell counts were performed using a
Neubauer chamber in a light-contrast microscope (Leica).
16S rRNA gene sequences were utilized to study the bacterial phylogeny and taxonomy
present in the sludge inoculum and in the reactor during phases I (adaption) and IV (OLR).
Briefly, the 16S rRNA amplicons of all samples were cloned into pGEMT-Easy vector and
then sequenced in an ABI 3100 automated sequencer (Applied biosystem) using a dye
terminator kit. The sequences were then used for phylogenic analysis. The experimental
details were described in Rampinelli et al. [10].
4.3. Results and discussion
4.3.1. Reactor start-up and biomass
Cell counts determined by the MPN technique were utilized as an indication of microbial
adaptation to the reactor conditions. This was performed only in the liquid phase inside the
FBR reactor (taken from port 1, figure 4.1), unlike in previous works [11]. This is because
any carrier (activated carbon) removal from the reactor would result in a decrease in the
sorbed biomass concentration. Therefore, specific rates were not determined, and bacterial
population was monitored only as a control parameter. The MPN figures (figure 4.2)
determined at the end of each phase, revealed an SRB population larger than 108MPN/mL,
93
which showed minor variations with both organic (OLR) and sulfate loadings (SLR) (table
4.1). In fact, during phase II, when the OLR was reduced from 12.34±0.98KgCOD/m3.d to
8.70±0.63KgCOD/m3.d, the SRB counts were reduced by almost one order of magnitude,
i.e., from 1.58x109MPN/mL (phase I) to 7.2x108MPN/mL (phase II). Similar behavior was
observed when the SLR was increased from 5.61±0.29KgSO42-/m3.d (phase III –
1.5x10 9MPN/mL) to 6.46±0.34KgSO42-/m3.d (phase IV – 1.1x109MPN/mL), but the
variations in cell counts were smaller. Similar bacterial counts were also determined by
Mizuno et al. [12], who studied sulfate reduction with sucrose in batch reactors, whereas
Bertolino et al. [13] found that the SRB population in the presence of lactate reached
SRB population (cell/mL, MPN)
9.5x10 9MPN/mL during sulfate reduction in a UASB reactor.
1E10
1E9
1E8
I
II
III
IV
V
Phases
Figure 4.2. Evolution of biomass concentration monitored on FBR reactor.
Molecular biology techniques [10] enabled a qualitative assessment of the microbial
diversity present in the FBR. The enrichment procedure successfully resulted in a diverse
SRB population, as shown in table 4.3, while inhibiting the growth of methanogens
because much microrganisms were not identified in the inoculum [13]. This microbial
94
diversity was expected due to the inoculum origin (domestic sewage treating reactor). In
addition, enrichment with Postgate C medium induced, as expected, the growth of
incomplete-oxidizing SRB (Desulfomonas, Desulfovibrio, Desulfolobus, Desulfobulbus
and Desulfotomaculum genera). The main metabolic pathways accounting for lactate and
glycerol oxidation as well as sulfate reduction by such microorganisms are discussed in
section 4.3.2.
Table 4.3.Microorganisms identified by molecular biology techniques in the inoculum
during FBR operation.
Microorganisms
11. Desulfomonas pigra (SF192152) (IO)
12. Desulfovibrio desulfuricans subsp. Desulfuricans str. ATCC
27774 (IO)
13. Desulfolobus sp. (IO)
14. Desulfovibrio vulgaris (IO)
15. Uncultured Desulfovibrio sp.Clone A37bac 16S ribosomal(IO)
16. Desulfobulbus sp. (EF442937) (PO)
17. Desulfobacter halotolerans DSM 11383 (NR026439) (AO)
18. Uncultured Desulfotomaculum sp.Clone BNB-488 (FJ898345)
(IO)
19. Methanogens
20. Clostridium sp.
IO – Incomplete – oxidizer BRS to acetate.
PO – Propionate – oxidizer BRS to acetate.
AO - Acetate – oxidizer BRS.
FBR
Start-up
Phases
inoculums
II toV
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
-
+
4.3.2. Reactor performance
Lactate as a substrate for sulfate reduction
A previous study [14] showed that a COD/sulfate ratio of 2.0 was optimum for SRB
growth under batch conditions (lactate), resulting in 98% sulfate reduction. Furthermore, a
95
study conducted under similar experimental conditions suggested lactate fermentation as
an important metabolic pathway during sulfate reduction in UASB reactors [15]. Because
Clostridium sp. was identified in the reactor inoculum (table 4.4), the COD/sulfate molar
ratio was set to 2.52 during phase I (table 4.2). Such a low ratio would avoid competition
between Clostridium sp. and SRB for lactate (because Clostridium sp. has a lower affinity
for lactate [16], i.e., sulfate reduction would not be limited by lactate depletion. In the
subsequent phases, both the lactate and sulfate concentrations were modified to determine
the optimum parameters for sulfate reduction. Such modifications as well as their effects
on the reactor performance are shown in figure 4.3.
96
(B )
(A )
10000
I
III
II
IV
100
V
COD removal (%)
COD (mg/L)
8000
IN
6000
5000
4000
III
70
60
50
40
30
2000
20
10
O UT
0
0
0
50
100
150
200
250
300
350
400
450
0
500
50
100
150
200
250
300
350
400
450
(C )
I
300 0
III
II
(D )
100
V
IV
IN
250 0
200 0
80 0
60 0
OUT
40 0
20 0
Sulfate reduction (%)
350 0
500
T im e (d ays)
T im e (d a ys )
Sulfate concentration
(mg/L)
V
IV
80
3000
1000
II
90
9000
7000
I
90
80
70
60
20
10
0
0
50
100
150
200
250
3 00
T im e (d a y s )
3 50
4 00
45 0
5 00
I
II
III
0
50
10 0
15 0
2 00
250
3 00
V
IV
0
350
400
4 50
500
T im e (d a ys )
Figure 4.3. Parameters monitored during sulfate reduction with lactate and glycerol. COD concentration (A) and removal efficiency (B). Sulfate
concentration (C) and removal efficiency (D). The first 50 days correspond to the adaptation period. The circle on figure 4.3D represents the
period when only glycerol was fed to the FBR.
97
First, the procedure selected for inoculum preparation and the steady increase in the organic
loading (lactate), shown in figure 4.3A, ensured fast biofilm formation, as suggested by the
reactor performance throughout the first fifty days, whereby sulfate removal yields increased
from 80 to 92% (figure 4.3D). In addition, the bacterial consortium present in the FBR
accounted for an average organic matter consumption that increased from 20% to 60% during
inoculation. Thereafter, the total organic matter consumption reached 53% (on average),
irrespective of the applied organic and sulfate loadings (phase I – 51%; II – 47%; III – 58%;
IV – 54%; figure 4.3B). These values are consistent with the work of Celis-Garcia et al.[17],
who reported a 60-70% overall COD consumption in a down-flow fluidized bed reactor
(DFFB).
Nevertheless, the electron flux for sulfate reduction actually varied with the OLR. For
instance, during phase I (12.34kgCOD/m3.d, table 4.1), for a minimum SRB population of
1.58 x 109MPN/mL (free cells – not attached to the activated carbon particles) (figure 4.2),
51% of the oxidized organic matter (figure 4.3B) was utilized for 97% sulfate reduction
(figure 4.3D). In phase II (8.70kgCOD/m3.d) (table 4.1), the SRB population showed better
utilization of the transferred electrons (64.7%). However, sulfate reduction decreased to 80%,
which was confirmed by a drop in the SRB population (to 7.20x108MPN/mL). In phase III
(11.82kgCOD/m3.d), the SRB counts increased again to 1.5x109MPN/mL, and 95% of inlet
sulfate was reduced, implying that 51.7% of the electrons were transferred to the anion (figure
4.3B). Phase IV again showed a decrease in sulfate-reduction performance (79%), derived
from an increase in SLR in the FBR to 6.46kgSO42-/m3.d (table 4.1), which indicates that
53.3% of the electrons were transferred to sulfate. These values were slightly worse than
those determined by Kaksonem et al. [18], who observed a 60-75% electron flux in an FBR
treating 1.49-3.3kgSO42-/m3.d. The worse performance during phases II and IV can be
98
justified by a decrease in the COD/sulfate ratio, which will be discussed later.
The stability of the FBR reactor was monitored by pH, VFA concentration and alkalinity
values (figure 4.4). The optimum pH for SRB growth is in the 7.0 - 7.8 range, although values
between 5 and 9 can be tolerated [19]. In the FBR, the pH value was defined by the balance
between VFA concentration and alkalinity. During incomplete lactate oxidation by SRB
(reaction 4, table 4.4), any acidity produced during acetate formation would be neutralized by
the alkalinity produced by the same metabolic pathway. Figure 4.4A shows that even with a
significant variation in VFA production (1000-3500mg/L), the produced alkalinity
(2500mg/L) was sufficient to maintain the system pH at values above 8 (figure 4.4B), with
the exception of a small period in phase IV (between 300 and 350 days) when a drop in
sulfate reduction resulted in lower alkalinity production. This high alkalinity concentration
may potentially be applied to treat mildly acid effluents, such as acid mine drainage, as
proposed by Nevatalo et al. [11].
Table 4.4. Anaerobic degradation reactions relevant to this study.
#
1
2
3
4
5
6
Chemical reaction
2lactate → 1.5acetate + H 3lactate → acetate + 2propionate + HCO
0 + H
Glycerol + 2H, O → acetate + HCO0 + 2H + 3H,
2lactate + SO,
+ → 2acetate + 2HCO0 + HS + H
Glycerol + 0.75SO,
+ → acetate + 0.75HS + HCO0
+1.25H + H, O
Glycerol + 1.25SO,
+ → 0.5acetate + 1.5H, CO0 + 0.5HCO0
+1.25HS + 0.75OH + 0.25H, O
∆G0’ (KJ)
-56.5
-169.7
-73.2
-160.1
-225.2
-424.5
99
10000
4000
Volatile Fatty Acids (mgHAc/L)
III
IV
V
8000
3000
BA
6000
2000
4000
1000
2000
VFA
0
0
50
100
150
200
250
300
350
400
450
(A)
Time (days)
9.0
-100
I
III
II
IV
V
8.5
-150
8.0
-200
7.5
pH
-250
7.0
-300
6.5
-350
6.0
-400
Eh
5.5
5.0
Eh (mV)
pH
Bicarbonate alkalinity (mgCaCO3/L)
II
I
-450
-500
50
100
150
200
250
300
350
Time (days)
400
450
(B)
Figure 4.4. Performance parameters of the FBR reactor, in different phases. The arrow in
figure 4.4A indicates a peak on VFA production.
Both the microbial characterization and the VFA profile suggest that sulfate reduction as well
as lactate oxidation predominantly followed reaction 4 (table 4.4) in which Desulfovibrio sp.
plays a key role. The VFA profile supports this finding because only acetate was detected in
the reactor effluent, although four other volatile fatty acids were also analyzed (propionate,
butyrate, isovalerate and isobutyrate). In addition, the predominance of incomplete-oxidizing
100
SRB (table 4.3) was also consistent with acetate concentrations produced from lactate
oxidation and sulfate reduction (table 4.5). Several studies have suggested acetate
accumulation during sulfidogenesis in fluidized bed reactors fed with lactate and ethanol
[17,20-21].
Table 4.5.Parameters of incomplete lactate oxidation reaction as a function of feed lactate and
sulfate (according to reaction 4).Values in mmol/L.
Phase
I
II
III
IV
Stoichiometrical ratio
Theoretical
Influent
Predicted*
Effluent (residual)
Influent
Predicted
Effluent (residual)
Influent
Predicted
Effluent (residual)
Influent
Predicted
Effluent (residual)
Consumed
Lactate
2
53 ± 2.9
40
0
38 ± 2.8
35
0
52 ± 2.6
48
0
52 ± 3.1
46
0
Consumed
sulfate
1
21 ± 0.85
20 ± 0.86**
0.67
22 ± 0.85
17 ± 2.50
4.79
25 ± 1.43
24 ± 2.03
1.71
29 ± 4.60
23 ± 3.81
6.04
Produced
acetate
2
0
40
42 ± 9.6
0
35
35 ± 8.2
0
48
46 ± 3.4
0
46
37 ± 5.3
*Predicted lactate and acetate concentrations were calculated from sulfate consumed
in FBR according to reaction 4.
** Consumed sulfate concentrations are actual values.
As predicted by reaction 4, in the reduction of 1mol/L of sulfate, 2mols/L of lactate was
oxidized. Therefore, higher lactate concentrations could enable the growth of microorganisms
that compete with SRB for lactate, whereas lactate concentrations below 2mol/L would limit
sulfate reduction. Such relationships can be observed by comparing phases I and II (table 4.1),
when the reactor operated with the same inlet sulfate concentration (21mmol/L sulfate).
During phase I, the lactate/sulfate molar ratio was set to 2.52±0.18, which reduced 97% of the
added sulfate (21±0.85mmol/L), producing 42±9.6mmol/L of acetate; this agrees fairly well
101
with the values predicted by reaction 4 (table 4.4). Conversely, when the lactate/sulfate molar
ratio was reduced to 1.73±.0.18 (phase II), sulfate reduction was reduced to 80% (figure
4.3D), implying that only 17±2.5mmol/L of sulfate was converted to 35±8.2mmol/L acetate
(table 4.5). Therefore, the lactate/sulfate molar ratio needed to be set to values above 2.0 to
ensure that there was no carbon source limitation for sulfate reduction. Values slightly above
2 are required because lactate is also degraded by other microorganisms found in the system,
such as Clostridium spp. During sulfate reduction in a UASB reactor treating lactate,
Bertolino et al. [13] reported a strong correlation between sulfate-reduction yields and the
organic loading rate (OLR), which was ascribed to a competition between acidogenic bacteria
and SRB. Similar results were observed when the effect of increasing sulfate loadings in the
reactor performance was assessed. Analyzing phases I, III and IV in which the organic
loading rate was held constant (~12KgCOD/m3.d), a similar SRB population in the reactor
(109cells/mL) was observed (figure 4.2), but because of the increase in the sulfate loading rate
(from 4.83±0.29KgSO42-/m3.d to 6.46±0.34KgSO42-/m3.d), a lower sulfate reduction
efficiency was achieved, as depicted in figure 4.3D, which is a consequence of the lower
COD/sulfate molar ratio.
As stated, for a lactate/sulfate molar ratio of 2.52 (phase I) in which lactate was not limiting,
there was almost 100% sulfate reduction and the attained residual anion concentration was
0.67mmol/L (64mg/L). When the same ratio was decreased to 2.08 (phase III), sulfate
reduction was slightly reduced to 95%, i.e., 24±2.03mmol/L of sulfate was reduced to
produce 46±3.4mmol/L of acetate (according to reaction 4, table 4.4). Furthermore, when the
lactate/sulfate molar ratio was reduced even further to 1.79 (phase IV), the sulfate reduction
efficiency was sharply reduced to 79%, which corresponds to 6.04mmol/L (580mg/L) sulfate
in the reactor effluent; therefore, lactate was limiting during phase IV (table 4.5).
102
Glycerol as the single organic substrate for sulfate reduction
The performance of the FBR was assessed in the presence of glycerol as an alternative carbon
and electron source (phases V). That was achieved by setting the glycerol concentration to
41mmol/L (the required amount to reduce 2.0g/L sulfate according to reaction 5, table 4.4),
whereas the lactate concentration was gradually reduced from 44mmol/L (phase IV) to zero
(table 4.2). This approach was carried out so that sulfate reduction would not be limited by the
carbon source, should the biomass be unable to metabolize glycerol.
During the period when lactate was progressively replaced by glycerol, sulfate reduction
varied from 80 to 92% to generate a COD removal efficiency of 58% (on average). This latter
figure is similar to that measured during sulfate reduction in the presence of lactate as the sole
carbon source (phases I-IV, figure 4.3B). Nevertheless, a change in the metabolic pathway
could be observed when glycerol was utilized as a single carbon source. Such a change was
observed in the profiles of the pH, Eh, alkalinity and VFA, particularly acetate (figure 4.4;
table 4.6).
Table 4.6. Parameters of incomplete glycerol and lactate oxidation reaction during the
beginning of phase V. Values in mmol/L.
Period
400-430
450-500
Influent
Glycerol
41
41
Lactate
44
0
Effluent
Sulfate
21
21
Sulfate
3.0
3.1
Acetate
59
9.0
Propionate
4.5
0
Butyrate
6.0
3.5
At the beginning of the substrate replacement process, when the CODglycerol/CODlactate mass
ratio was 1:1 (from days 402 to 422), a high organic loading rate (ORL=15.21Kg/m3.d) (table
4.2) resulted in high acetate production (59mmol/L, table 4.6), with minor concentrations of
propionate (4.5mmol/L) and butyrate (6.0mmol/L). This high VFA content induced a
103
decrease in the pH inside the reactor (figure 4.4B). Both phenomena could be explained by
the simultaneous growth of both SRB and acidogenic microorganisms (reactions 1, 2 and 3,
table 4.4), which is consistent with the results of other studies [15] and explained the higher
solution potential observed during this period (-350 mV, figure 4.4B). It is worth nothing that
there was also a slight drop in sulfate reduction yields (figure 4.3D). From the 422nd day
onward, when the ORL was reduced to 6.64Kg/m3.d, the sharp reduction in alkalinity (figure
4.4A) suggested that the system was under stress because there was a peak in the VFA
production (arrow in figure 4.4A), which resulted in a drop in pH (from 8.0 to 6.5). This
phenomenon indicated a change in the metabolic pathway. This was because glycerol
oxidation during sulfate reduction (reactions 5 or 6, table 4.4) produced less alkalinity than
that generated during lactate degradation (reaction 4, table 4.4). The reactor performance
approached a steady state when 70% of the inlet COD was due to glycerol (day 445;
ORL=5.61Kg/m3.d) until the end of the experiment (figures 4.3D and 4.5A). Under such
operational conditions, the alkalinity and VFA profiles reached 0.5g/L and 1.0g/L,
respectively, whereas the solution potential was reduced to its lowest value (-450 mV) and the
reactor pH stabilized again at values above 7.0.
As glycerol became the sole carbon and electron source (day 445), the 18mmol/L reduction in
sulfate (on average) produced only 9mmol/L of acetate, i.e., a “produced acetate”/“reduced
sulfate” ratio of 0.5 was achieved. Although this ratio is lower than the value of 1.33 observed
for a pure Desulfovibrio strain (reaction 5, table 4.4) [3], it is closer to the value determined
by Dinkel et al.[4], who hypothesized a metabolic pathway for glycerol degradation. Such a
hypothesis proposed a syntrophic cooperation between acetogenic bacteria and SRB, which
could be chemically represented by reaction 6 (table 4.4), when the “produced
acetate”/“reduced sulfate” molar ratio was 0.4. Given the microbial diversity determined in
104
the FBR (table 4.3), reaction 6 explained the findings of the present work because the actual
values of alkalinity and acetate concentrations could be predicted by this reaction.
Batch experiments were performed to gain more insight into the glycerol degradation
pathway. Figure 4.5A depicts the profiles of glycerol, sulfate and pH, whereas figure 4.5B
presents the concentration of selected VFA and total organic carbon (TOC) assays. Both
figures suggest two phases during sulfate reduction in the presence of glycerol. During the
first 13h (phase α), glycerol was almost fully degraded, which resulted in acetate and butyrate
accumulation, along with a pH reduction from 8.0 to 5.5, for a negligible sulfate reduction. A
comparison between the theoretical TOC, determined from the carbon present in the detected
VFA, and the experimentally analyzed TOC (figure 4.5B) suggested the presence of one or
more intermediary products in addition to those two species. It has been shown that
Clostridium sp.(table 4.3) is able to ferment glycerol, producing compounds such as 1,3-PD
[23-25] (a major species not analyzed), acetate and butyrate [24-25], both of which were
detected in the present work (figure 4.5B). Therefore, the formation of 1,3-PD was suggested,
which is supported by the concomitant formation of butyrate and acetate in the reactor.
1.0
18
8.0
0.8
16
0.7
14
0.6
12
pH
0.5
10
0.4
8
0.3
[SO4 ]
0.2
6
-1
2-
4
20
40
60
80
Time (hours)
100
120
7.0
6.5
6.0
5.5
5.0
0.1
0
7.5
pH
Absorbance (660nm)
OD
8.5
Sulfate concentration (mmol.L )
0.9
20
140
(A)
105
48
100
measured TOC
-1
(mmol.L )
60
Analytical TOC
40
30
40
Glycerol
acetate
Butyrate
10
5.0
-1
7.5
20
Total Organic Carbon
80
44
(mmol.L )
Glycerol and VFA concentration
120
2.5
0
0.0
0
20
40
60
80
100
Tim e (hours)
120
(B)
Figure 4.5. Parameters monitored in the batch experiment.(A) bacterial growth (OD), sulfate
concentration and pH. (B) VFA profile and total organic carbon.
Because 1,3-PD is readily degraded by the Desulfovibrio [3]; it is thus hypothesized that only
after glycerol was converted to 1,3-PD was sulfate reduction initiated (phase β). Sulfate
reduction produced alkalinity (reaction 5, table 4.4), whereby the pH increased from 5.5 to
6.5. Meanwhile, the analyzed TOC concentrations approached their theoretical values during
phase β (figure 4.5B), which were equivalent to the sum of the acetate and butyrate
concentrations (up to the end of the experiment), suggesting that the intermediary product
(likely 1,3-PD) was progressively converted to the two VFA.
To summarize, it is proposed that glycerol itself was not utilized for sulfate reduction; instead,
it first needed to be converted to an intermediary product, which is believed to be 1,3-PD,
before being utilized by SRB [3]. Glycerol degradation would account for the production of
19mmol/L of butyrate and 17mmol/L of acetate during the reduction of 13.5mmol/L of
sulfate. Therefore, the metabolic pathway for sulfate reduction with glycerol may occur in two
106
steps: (i) glycerol fermentation to butyrate and 1,3-PD by Clostridium spp., followed by (ii)
1,3-PD degradation by Desulfovibrio spp. or other SRB (table 4.3). Such a metabolic pathway
is detailed in figure 4.6 along with that of lactate [15].
Figure 4.6. Metabolic pathways for glycerol and lactate degradation during sulfate reduction.
X is an electron carrier.
The main shortcoming of anaerobic sulfate reduction is the cost of the carbon source and the
formation of by-products such as acetate and butyrate, which increases the residual COD [5].
Sulfate reduction with glycerol as the carbon source produced a reactor effluent with lower
acetate and butyrate concentrations than those observed when lactate was the single carbon
source. During phase I, when the highest sulfate reduction (97%) was observed, the low COD
consumption (50%) was derived from the incomplete lactate oxidation and resulted in a
reactor effluent with a high acetate content (2480mg/L; table 4.5) and thus a high residual
COD (approximately 2500mgCOD/L, on average). Sulfate reduction in the presence of
glycerol was lower (89%) and produced a residual sulfate concentration of 254mg/L, on
107
average. Nevertheless, 75% of the influent COD was consumed, which represents a residual
COD 2.5 times lower than that achieved when lactate was the carbon source. Future work will
focus on the use of crude glycerol (g-phase) as the carbon source and on the analysis of 1,3PD during sulfate reduction. G-phase would be an interesting alternative for a cost-effective
implementation of sulfate-reducing bioreactors and a promising solution for the large-scale
production of crude glycerol by the biodiesel industry.
4.4. Conclusions
Glycerol can be utilized as a carbon source for sulfate reduction in fluidized bed reactors in
the presence of a mixed bacterial population containing at least sulfate-reducing bacteria and
fermentative microorganisms. Compared with lactate, the standard organic and electron
source for sulfate-reducing bacteria, the results revealed similar performances. Lactate
showed slightly better sulfate reduction (97%) compared to glycerol (89%) but also a higher
residual chemical oxygen demand (2500mg/L). Conversely, glycerol produced an effluent
COD of approximately 1000mg/L. The glycerol oxidation mechanism involves first
fermentation to an intermediary product, which is believed to be 1,3-propanediol; this
intermediary is subsequently oxidized by sulfate-reducing bacteria to butyrate and acetate.
Because lactate is not a cost-effective carbon source, glycerol could be an alternative to
ethanol, which is widely applied industrially. As observed with lactate, the residual COD is
lower when glycerol is utilized.
108
The financial support from the funding agencies FINEP, FAPEMIG, CNPq and CAPES is
gratefully appreciated. The Conselho Nacional de Pesquisas - CNPq scholarships to S. M.
Bertolino, L. A. Melgaço, S. F. Aquino and V. A. Leão are especially acknowledged.
4.6. References
[1] N. Kolesárová, M. Hutnan, I. Bodík, V. Spalková, Utilization of Biodiesel By-Products
for Biogas Production, Journal of Biomedicine and Biotechnology, 2011 (2011) 1-15.
[2] Y. Yang, K. Tsukahara, S. Sawayama, Biodegradation and methane production from
glycerol-contaning synthetic waste with fixed-bed bioreactor under mesophilic and
thermophilic anaerobic conditions Process Biochemistry, 43 (2008) 362-367.
[3] A.I. Qatibi, Fermentation du lactate, du glycerol et des diols par les bactéris sulfatoréductrices du genere Desulfovibrio, in, 1990.
[4] V. Dinkel, F. Frechen, A. Dinkel, Y. Smirnov, S. Kalyuzhnyi, Kinetics of anaerobic
biodegradation of glycerol by sulfate-reducing bacteria, Applied Biochemistry and
Microbiology, 46 (2010) 712-718.
[5] INAP, Treatment of sulphate in mine effluentes, in, International network for acid
prevention, 2003, pp. 129.
[6] S.F. Aquino, A.Y. Hu, A. Akram, D.C. Stuckey, Characterization of dissolved compounds
in submerged anaerobic membrane bioreactors (SAMBRs), Journal of Chemical Technology
& Biotechnology, 81 (2006) 1894-1904.
[7] APHA, Standard Methods for the examination of water and wastewater, Washington, DC,
2005.
[8] P. Bondioli, L.D. Bella, An alternative spectrophotometric method for the determination
of free glycerol in biodiesel,European Journal of Lipid Science and Technology, 107 (2005)
153-157.
[9] J.R. Postgate, Versatile medium for the enumeration of sulfate-reducing bacteria., Applied
Microbiology, 11 (1963) 265-267.
109
[10] L. Rampinelli, R. Azevedo, M. Teixeira, R. Guerra-Sá, V. Leão, A sulfate-reducing
bacterium with unusual growing capacity in moderately acidic conditions, Biodegradation, 19
(2008) 613-619.
[11] L.M. Nevatalo, A.E. Mäkinen, A.H. Kaksonen, J.A. Puhakka, Biological hydrogen
sulfide production in an ethanol lactate fed fluidized-bed bioreactor, Bioresource Technology,
101 (2010) 276-284.
[12] O. Mizuno, Y.Y. Li, T. Noike, The behavior of sulfate-reducing bacteria in acidogenic
phase of anaerobic digestion, Water Research, 32 (1998) 1626-1634.
[13] S.M. Bertolino, I.C.B. Rodrigues, R. Guerra-Sá, S.F. Aquino, V.A. Leão, Implications of
volatile fatty acid profile on the metabolic pathway during continuous sulfate reduction,
Journal of Environmental Management, 103 (2012) 15-23.
[14] L.P. Barbosa, S.M. Bertolino, P.C. Freitas, V.A. Oliveira, P.D. Pina, V.A. Leão, M.C.
Teixeira, Effects of differnt COD/sulfate ratios on the growth of metal tolerant sulfate
reducing bacteria (SRB), biohydrometallurgy: a meeting point between microbial ecology,
metal recovery process and environmental remediation., in:
Proceedings of
Biohydrometallurgy, Bariloche, 2009, pp. 569-572.
[15] Y. Zhao, N. Ren, A. Wang, Contributions of fermentative acidogenic bacteria and
sulfate-reducing bacteria to lactate degradation and sulfate reduction, Chemosphere, 72
(2008) 233-242.
[16] G. Zellner, F. Neudorfer, H. Diekmann, Degradation of Lactate by an Anaerobic Mixed
Culture in a Fluidized-Bed Reactor, Water Research, 28 (1994) 1337-1340.
[17] L.B. Celis-García, E. Razo-Flores, O. Monroy, Performance of a down-flow fluidized
bed reactor under sulfate reduction conditions using volatile fatty acids as electron donors,
Biotechnology and Bioengineering, 97 (2007) 771-779.
[18] A. Kaksonen, P. Franzmann, J. Puhakka, Performance and Ethanol Oxidation Kinetics of
a Sulfate-Reducing Fluidized-Bed Reactor Treating Acidic Metal-Containing Wastewater,
Biodegradation, 14 (2003) 207-217.
[19] L.L. Barton, Sulfate-reducing bacteria, Plenum Press, New York, 1995.
[20] E. Sahinkaya, B. Özkaya, A.H. Kaksonen, J.A. Puhakka, Sulfidogenic fluidized-bed
treatment of metal-containing wastewater at low and high temperatures, Biotechnology and
Bioengineering, 96 (2007) 1064-1072.
[21] S. Nagpal, S. Chuichulcherm, L. Peeva, A. Livingston, Microbial sulfate reduction in a
liquid–solid fluidized bed reactor, Biotechnology and Bioengineering, 70 (2000) 370-380.
110
[22] M. González-Pajuelo, J. Andrade, I. Vasconcelos, Production of 1,3-Propanediol by
Clostridium butyricum; VPI 3266 in continuous cultures with high yield and productivity,
Journal of Industrial Microbiology & Biotechnology, 32 (2005) 391-396.
[23] K. Leja, K. Czaczyk, K. Myszka, Biotecnhological synthesis of 1,3-propanediol using
Clostridium ssp., African Journal of Biotechnology 10 (2011) 11093-11101.
[24] S. Papanikolaou, P. Ruiz-Sanchez, B. Pariset, F. Blanchard, M. Fick, High production of
1,3-propanediol from industrial glycerol by a newly isolated Clostridium butyricum strain,
Journal of Biotechnology, 77 (2000) 191-208.
[25] A. Drozdzynska, K. Leja, K. Czaczyk, Biotechnological production of 1,3-propanediol
from crude glycerol, Journal of Biotechonology 92 (2011) 92-100.
111
CAPÍTULO 5
CONSIDERAÇÕES FINAIS
O objetivo deste trabalho foi avaliar o processo anaeróbio sulfetogênico em dois reatores
contínuos (UASB e Leito Fluidizado) e também a aplicação de uma fonte alternativa de
carbono e elétrons para o tratamento de efluentes com alta carga de sulfato. Baseado no
estudo desenvolvido, as seguintes considerações podem ser feitas:
•
O reator de leito suspenso (UASB) necessitou de um longo tempo (superior a 300
dias) para alcançar um valor ideal para a população de BRS (>108MPN/mL). Além
disso, uma baixa velocidade ascendente (etapas sem recirculação) repercutiu em baixa
eficiência de remoção de sulfato que atingiu apenas o valor máximo de 66%. Em
contraste, o reator de Leito Fluidizado, no qual a biomassa estava retidano carvão
ativado e que opera com alta velocidade ascendente, levou apenas 45 dias para atingir
eficiências de remoção superiores a 90%.
•
O uso de lactato com fonte de carbono e elétrons favoreceu a formação de uma
biomassa predominantemente de bactérias redutoras de sulfato que oxidam
incompletamente o substrato a acetato (representadas, por exemplo, pelos gêneros
Desulfovibiro spp. e Desulfomonas spp.), além de bactérias fermentativas
(Clostridiumspp.). O acúmulo de acetato (>2000mg/L) indicou ausência de atividade
acetoclástica na biomassa dos dois reatores.
•
O melhor desempenho do reator UASB foi atingido com a recirculação do efluente
devido à melhora nas condições de transferência de massa. Para uma carga de sulfato
112
aplicada de 2,0g/L.d, o reator removeu em média 1,6g/L.d de sulfato. Os resultados
indicam que, para um reator com biomassa suspensa e sem recirculação, o processo
sulfetogênico foi prejudicado pela competição, pelo substrato,entre bactérias redutoras
de sulfato e fermentativas uma vez que, a espécie propionato (indicativo de
fermentação) foi observada em todas as fases, principalmente naquelas que operaram
com alta carga orgânica (>5.0gDQO/L.d). Por outro lado, durante o monitoramento do
reator de leito fluidizado, degradando lactato,entre os ácidos graxos monitorados
(acetato, propionato, butirato, isobutirato e isovalérato) apenas o acetato foi detectado
no efluente. E este é proveniente da atividade das bactérias redutoras de sulfato.
•
No reator de leito fluidizado, em virtude das elevadas eficiências de redução de
sulfato, praticamente toda a geração de elétrons, proveniente da oxidação do lactato,
foi canalizada para a redução do sulfato a sulfeto, pelas BRS. Para uma carga de
sulfato aplicada de 4,82g/L.d, uma taxa de redução de 4,67gSO42-/L.d foi alcançada na
presença de lactato. Para as fases com razões DQO/sulfato superiores a 2,0, o
desempenho do RLF produziu um efluente com concentração de sulfato inferior a
250mg/L.
•
O biofilme formado no reator de leito fluidizado, com o predomínio das BRS do
gênero Desulfovibrio e das fermentativas do gênero Clostridium, foi capaz de utilizar
o glicerol como única fonte de carbono e elétron para a redução de 89% do sulfato.A
taxa específica de redução do ânion, com o glicerol (0.172±0.010gSO42-/(gSSV.d) foi
similar àquela observada com o lactato, ou seja, 0.191±0.016gSO42-/(gSSV.d).Este
resultado indica o potencial uso do glicerol brutoproveniente da produção do biodiesel
113
para o tratamento de efluentes industriais contendo sulfato em reatores anaeróbios
sulfetogênicos.
Pode-se concluir que a redução de sulfato no reator UASB, sem recirculação não é favorecida,
o que pode ser justificado pela: (i) ausência de granulação, ou seja, foi formado um lodo de
baixa densidade, facilmente arrastado com o efluente; (ii) baixa velocidade ascensional (baixa
vazão e, portanto, elevados tempos de residência), necessária para evitar o arraste da
biomassa. A elevação da relação DQO/sulfato pode não resultar no aumento da redução de
sulfato, uma vez que atividade fermentativa é favorecida. A recirculação da biomassa permite
o aumento da velocidade ascensional e consequentemente, melhora das condições de mistura,
favorecendo a atividade sulfetogênica em detrimento da fermentativa.
No reator de leito fluidizado, condições ideais de transferência de massa e a retenção da
biomassa no carvão ativado permitiram elevadas remoções de sulfato. A principal variável foi
a relação DQO/sulfato, que por sua vez foi definida pela diversidade de BRS presente no
reator. O glicerol pode ser utilizado como fonte de carbono, entretanto, necessita-se atividade
fermentativa para sua plena utilização no processo.
114
CONTRIBUIÇÕES AO CONHECIMENTO
No contexto do tratamento de esgotos domésticos em reatores UASB, diversos estudos
relatam a redução do sulfato pela BRS somente na última etapa de degradação da matéria
orgânica, ou seja, em competição com as arqueias metanogênicas. Essa tese correlacionou
pela primeira vez a atividade sulfetogênica com a fermentativa, nesses reatores, para elevadas
cargas de sulfato. Além disso, todo desempenho dos reatores pode ser discutido com base
apenas na atividade destes dois grupos de micro-organismos, ou seja, as BRS que oxidam
incompletamente o lactato e as fermentativas (BF). O predomínio do grupo de BRS
acetoclásticasdeterminou a melhor razão COD/sulfato, enquanto a condição hidrodinâmica do
reator definiu o grau de competição entre as BRS e as BF.O trabalho mostrou ainda que a
elevação da carga orgânica (razão DQO/sulfato > 2.5) não implica em melhora na redução de
sulfato (com lactato), uma vez que uma alta carga orgânica induz a atividade fermentativa.
Isso ocorre por que as BF possuem maior taxa de crescimento e menor afinidade pelo lactato,
como substrato.
Independentemente do tipo de reator, a seleção de um substrato orgânico adequado para o
cultivo de BRS tem dificultado a implantação industrial desta biotecnologia. A literatura
sugere que as BRS são capazes de oxidar diretamente o glicerol. Entretanto, foi observado,
para culturas mistas de BRS e BF, a rápida fermentação deste antes do início da redução do
sulfato e foi proposto que o glicerol precisa ser convertido a um produto intermediário (tal
como o 1,3-Propanodiol, a ser confirmado) para então ser oxidado pelas BRS, durante a
redução de sulfato. Foi também demonstrada pela primeira vez, a redução de sulfato tendo o
glicerol com fonte de carbono e elétrons, em reatores de leito fluidizado. Isso possibilita o uso
115
da g-fase, sub-produto da produção do biodiesel (glicerol bruto),na biotecnologia de redução
do sulfato. Tem-se dessa forma um substrato orgânico mais econômico, e mais uma solução
e/ou aplicação para estesub-produtoda produção do biodiesel.
116
CAPÍTULO 7
SUGESTÕES PARA TRABALHOS FUTUROS
Diante dos desafios encontrados durante o período de elaboração e dos resultados obtidos
nesta tese, são sugeridos alguns temas relevantes a serem estudados futuramente, relacionados
ao tratamento de efluentes contendo sulfato e metais:
i) Avaliar a possibilidade de utilizar a alta alcalinidade produzida no reator UASB para
a precipitação de metais como cobre, níquel emanganês.
ii) Investigar o uso de glicerol bruto proveniente da produção de biodiesel como única
fonte de carbono e elétrons para a redução de sulfato.
iii) Conhecero desempenho dos reatores UASB e de Leito Fluidizado, com outras fontes
de carbono, preferencialmente sub-produtos industriais.
iv) Estudar um sistema completo de tratamento de efluentes ácidos contendo sulfato
emetais, a partir do reator de Leito Fluidizado.
117
ANEXO
ARTIGO PUBLICADO NO PERIÓDICO:
“Journal of Environmental management”
118
119

Tese de doutorado - REPOSITORIO INSTITUCIONAL DA UFOP

Transcription

Similar documents

Quiz 5 Key

Nutra-JOINT Brochure

Data Sheet-Gentamicin 2015

alkoxylation - Buss ChemTech

PoaCure Applications at Callippe Preserve Golf Course

28/25 Jersey Optimum

Keuzes voor electriciteitsopwekking