You are on page 1of 13

Unraveling the Influence of the COD/sulfate Ratio on the

Organic Matter Removal and Methane Production from the


Biodigestion of Sugarcane Vinasse

L. S. M. Kiyuna*, L. T. Fuess* and M. Zaiat*

* Biological Processes Laboratory, São Carlos School of Engineering, University of São Paulo, 1100 João
Dagnone Avenue, 13563-120, São Carlos, SP, Brazil
(E-mail: luma.kiyuna@gmail.com; ltfuess@sc.usp.br; zaiat@sc.usp.br)

Abstract
Sulfate accumulation in vinasse throughout the sugarcane harvest is a common characteristic in
sugar and ethanol plants. However, little is known on the influence of sulfate over the anaerobic
digestion (AD) of vinasse. The biodigestion of sulfate-rich wastewaters stimulates the sulfidogenic
process, so that sulfate reducing bacteria may outcompete methanogens by common substrates.
Inhibitory effects from sulfide may also be observed in such conditions. In this context, this paper
assessed the influence of different COD/sulfate ratios (12.0, 10.0 and 7.5) on both COD removal
and methane production from sugarcane vinasse AD. Batch assays were conducted in thermophilic
conditions. At a COD/sulfate ratio of 7.5, methane production was 35% lower compared to a ratio
of 12.0, considering the deviation of about 13.6% of the electron flow to sulfidogenesis. COD and
sulfate removal profiles indicated that deviation of electrons to sulfidogenesis were negligible at
COD/sulfate ratios higher than 25, considering an exponential increase in methane production. At
a COD/sulfate ratio of 12.0, less than 7.6% of the electron flow was diverted to sulfidogenesis,
only slightly impairing methane production. Organic matter degradation was not greatly affected
by sulfidogenesis, with COD removal levels higher than 80%, regardless of the initial COD/sulfate
ratio.

Keywords
Sugarcane biorefinery; bioenergy recovery; sulfate reduction; kinetic assessment

INTRODUCTION
The application of AD to vinasse, the main wastewater from ethanol production, has been widely
studied, considering the suitability of this wastewater to methane production and bioenergy
recovery. Several studies considered the use of different operational conditions and reactor
configurations, e.g. upflow sludge blanket and fixed-film reactors; sequencing batch reactors;
fluidized-bed reactors, among others, to vinasses resulting from different feedstocks, such as
sugarcane, beet, corn, cassava and cellulosic materials (Wilkie et al., 2000). Organic matter removal
efficiencies and methane yields respectively as high as 80% and 0.31 Nm3CH4 kg-1COD (Kumar et
al., 2007) were achieved in such studies.

Despite the successful results previously reported, the presence of specific compounds in vinasse,
such as recalcitrant (melanoidins and phenols) and interfering (mainly sulfate) compounds (Ferreira
et al., 2011; Fuess and Garcia, 2015), must be carefully considered based on the potential inhibition
of the anaerobic microbial populations, especially methanogenic archaea (MA). Among such
compounds, the presence of average-to-high levels of sulfate in vinasses from sugarcane is
highlighted in this study. Sugarcane-based distilleries widely employ sulfuric acid to prevent
microbial contamination and yeast flocculation in fermentation vessels, leading to sulfate
concentrations as high as 6.4 g L-1 in vinasse (Costa et al., 1986). The use of molasses, a residual
carbohydrate-rich solution from sugar plants, in the production of ethanol tends to enhance the
levels of sulfate in vinasse, as sulfuric acid is also employed in the step of sugar clarification (Fuess
and Garcia, 2015).

The application of sulfate-rich wastewaters to AD systems stimulates the sulfidogenesis process, in


which the sulfate is reduced mainly to sulfide by a specific group of microorganisms, the sulfate
reducing bacteria (SRB) (Lens et al., 1998; Vilela et al., 2014). SRB compete with MA by common
substrates, such as acetate and hydrogen (Vilela et al., 2014), limiting the extraction of energy
through methane. The toxicity of sulfide may also suppress the methanogenic activity in AD
systems, either by directly permeating the cell membrane as the non-ionized form (H2S) and
denaturing specific proteins or by indirectly enhancing the precipitation of essential metals, as the
ionized forms (HS- and S2-) (Vela et al., 2002; Chen et al., 2008). On the other hand, at conditions
of sulfate shortage established populations of SRB may act as acetogenic bacteria (Vilela et al.,
2014), partially oxidizing organic acids to acetate and consequently enhancing the activity of MA in
a syntrophic association (Vela et al., 2002).

Although the literature presents several studies on the influence of the sulfate reducing process on
methanogenic systems, investigations are usually conducted using synthetic wastewaters (Vela et
al., 2002; Camiloti et al., 2014; Vilela et al., 2014). Such trend indicates that the role of
sulfidogenesis in anaerobic reactors applied to real wastewaters, especially sugarcane vinasse, still
requires further investigations, in an effort to highlight whether conditions of inhibition or
performance enhancement should be observed. Using the COD/sulfate ratio as a reference
parameter, performance losses on both methane production and organic matter removal should not
be observed for values higher than 10 (Lens et al., 1998), since sulfide concentrations below the
critical levels to MA – 150 mg L-1 – should be expected in the reactors. The COD/sulfate ratio in
sugarcane vinasse usually reaches values higher than 15, however, in concentrated samples this
ratio may reach values close to 5, especially at the end of the sugarcane harvest.

In this context, this study investigated the influence of the COD/sulfate ratio on the AD of
sugarcane vinasse, focusing on conditions close to the potentially limiting COD/sulfate ratio value,
i.e., 10. Different COD/sulfate ratio values were assessed in batch assays, so that kinetic parameters
were obtained for organic matter and sulfate removal, as well as methane production. It should be
highlighted that this study complements a gap regarding the AD of sulfate-rich vinasses, providing
an understanding on how the sulfate reduction may affect the methane production from such
wastewater.

METHODS

Sugarcane vinasse
Vinasse samples were collected from a full-scale ethanol and sugar-plant located at Pradópolis, SP,
Brazil. The sampling was carried out at the beginning of the sugarcane harvest, in an attempt to
prevent the accumulation of high sulfate concentrations. The sample was filtered with a 3 μm paper
filter (Nalgon, density of 80 g m-2) to remove excessive concentrations of suspended solids. The
main compositional characteristics of the filtered vinasse included: total chemical oxygen demand
(TCOD) = 24.2 g L-1, soluble chemical oxygen demand (SCOD) = 21.6 g L-1, biochemical oxygen
demand (BOD) = 19.2 g L-1, volatile solids (VS) = 18.6 g L-1, volatile suspended solids (VSS) =
1.53 g L-1, total phosphorus = 147 mg L-1, total Kjeldahl nitrogen (TKN) = 774 mg L-1, sulfate
(SO42-) = 1.8 g L-1 and pH = 4.6.

Microbial consortium
The inoculum (0.186 gVS g-1TS) was obtained from a full-scale thermophilic (55°C) upflow
anaerobic sludge blanket reactor applied to the biodigestion of sugarcane vinasse. The non-granular
methanogenic sludge was sampled directly from the blanket zone of the reactor.
Experimental procedure
The batch reactors consisted of 500 mL glass flasks (DuranTM flasks), so that each one was filled
with 300 mL of sugarcane vinasse. The pH of vinasse was adjusted to 7.0 by adding sodium
bicarbonate (NaHCO3) prior to filling the flasks.

Three COD/sulfate ratio values were investigated: 12.0, 10.0 and 7.5, in terms of the SCOD. The
COD/sulfate ratio of 12.0 corresponded to the raw vinasse, whereas the other values for the ratio
were obtained by adding sodium sulfate (Na2SO4). For each COD/sulfate ratio six flasks were
analyzed: three for the gaseous phase, i.e., biogas composition, and three for the liquid phase, i.e.,
COD, sulfate and volatile fatty acids (VFA).

An inoculum/substrate ratio of 0.3 gVS g-1SCOD was initially applied to each flask. After filling
the flasks with the vinasse and the sludge, nitrogen was fluxed for 5 min to guarantee an anaerobic
environment in the reactors. The flasks were then placed in a shaker (430, Nova Ética, Vargem
Grande Paulista, SP, Brazil), with temperature and agitation conditions set as 55°C and 120 rpm,
respectively. Gas and liquid samples were periodically collected from the corresponding flasks.
Prior to the sampling, the internal pressure was measured in each flask using a pressure gauge
(MDD50, T&S Equipamentos Eletrônicos, São Carlos, SP, Brazil). Less than 10% of the overall
bulk liquid volume was taken for analyses throughout the experiment.

Analytical methods
TCOD, SCOD, sulfate, VFA and biogas composition were periodically analyzed during the
experimental run. At the end of the experiment, additional analyses regarding the partial alkalinity
(PA) and total dissolved sulfide (TDS) were carried out. Sulfide species in solution, i.e.,
undissociated and dissociated sulfide, were calculated using TDS, pKa values and the pH measured
in each flask. The values for TCOD, SCOD, VSS and TDS were obtained according to the
procedures described in the Standard Methods for the Examination of Water and Wastewater. Prior
to SCOD analyses zinc sulfate (ZnSO4) was added in excess to the samples to remove the
interference of sulfide in the measurements. The concentrations of sulfate were determined by ionic
chromatography (Dionex ICS-5000, Thermo Fisher Scientific Inc., Sunnyvale, CA, USA), using an
ion suppressor and a conductivity detector, as described by Costa and Foresti (2014). VFA and
solvents were analyzed by gas chromatography (GC-2010, Shimadzu Scientific Instruments,
Columbia, MD, USA) using a flame ionization detector (FID). Details on the sample preparation
and analytical conditions may be obtained in Adorno et al. (2014). The biogas composition
(methane – CH4 and carbon dioxide – CO2) was also determined by gas chromatography (GC-2014,
Shimadzu Scientific Instruments, Columbia, MD, USA), using a thermal conductivity detector
(TCD), according to Lebrero et al. (2015). Finally, PA was analyzed according to the protocol
described by Ripley et al., 1986).

Data analysis and calculations: electron flow and kinetic parameters


Electron flows to methanogenesis and sulfidogenesis were calculated as proposed by Vilela et al.
(2014), considering that SRB oxidize 0.67 g of organic matter (as SCOD) to reduce each gram of
sulfate (Lens et al., 1998). Equation (1) shows the determination of the electron flow considering
the total SCOD removal (ERSCOD, in %), whereas Equations (2) and (3) determine respectively
electron flows associated with SCOD removal via sulfidogenesis (ERSCODSO4, in %) and
methanogenesis (ERSCODCH4, in %).
In Equations (1) to (3) SCODO and SCODR are respectively the initial and residual concentrations
of organic matter measured in vinasse, in g L-1; whereas SO4O and SO4R are respectively the initial
and residual concentrations of sulfate measured in vinasse, in g L-1. Calculations assumed that the
total amount of sulfate initially applied to each flask was reduced to sulfide.

A modified Gompertz model (Zwietering et al., 1990) was fitted to the cumulative methane
production experimental data to obtain the methane production potential (PM, in L), methane
production rate (RM, in L d-1) and lag phase (λ, in d), according to Equation (4).

In Equation (4) VM is the cumulative methane production, in L (VM = PM when t → ∞); t is the
incubation time, in days; and e is the Euler’s number (2.71828).

The parameters PM, RM and λ were obtained using the software OriginPro 8 (OriginLab
Corporation, Northampton, MA, USA) by employing the Levenberg-Marquardt algorithm. The
specific methane production potential (PM’ in L kg-1VS) and specific methane production rate (RM’,
in L kg-1VS day-1) were obtained respectively by dividing PM and RM by the amount of
methanogenic sludge (kgVS) added to the flasks.

A first-order kinetic model with residual organic matter concentration (Camiloti et al., 2014) was
fitted to the temporal profiles of SCOD variation, according to Equation (5). The first-order kinetic
coefficient for SCOD (k1SCOD, day-1) was obtained for each COD/sulfate ratio, so that its estimation
was also based on the use of the software OriginPro 8 (OriginLab Corporation, Northampton, MA,
USA), with the Levenberg-Marquardt algorithm. The specific first-order kinetic coefficient
(k1SCOD’, kg-1VS day-1) was obtained by dividing k1SCOD by the amount of methanogenic sludge
(kgVS) added to the flasks.

In Equation (5) COSCOD and CRSCOD correspond respectively to the initial (t = 0) and residual
concentrations of organic matter (as SCOD) mathematically fitted, in g L-1; and t is the incubation
time, in days.

With respect to sulfate, a first-order kinetic model without residual concentration – Equation (6) –
was fitted to the experimental data. Estimation procedures were similar to the one above described
for SCOD, so that the first-order kinetic coefficient for sulfate (k1SO4, day-1) was obtained for each
COD/sulfate ratio. The specific first-order kinetic coefficient (k1SO4’, kg-1VS day-1) was also
obtained by dividing k1SO4 by the amount of sludge (kgVS) added to the flasks.

In Equation (6) COSO4 is the initial (t = 0) concentration of sulfate, in g L-1; and t is the incubation
time, in days.

RESULTS AND DISCUSSION

Methane production
Cumulative methane production profiles for both experimental data and fitted models are depicted
in Figures 1a-c, whereas the parameters mathematically obtained are summarized in Table 1. The
results indicate that the lower the COD/sulfate ratio, i.e., the higher the sulfate concentration, the
lower the cumulative methane production, which suggests that the methanogenic activity was
negatively impacted by sulfidogenesis. Regardless of the COD/sulfate ratio, methane production
showed two distinct patterns, considering the establishment of an initial plateau (first fragments,
after approximately 20, 40 and 60 days of incubation, respectively for the ratios of 12.0, 10.0 and
7.5) and then a marked increase up to the establishment of a second plateau (second fragments,
Figures 1a-c). This pattern may be attributed to the deviation of the electron flow (Figure 2d-f) to
the organic matter conversion via sulfidogenesis in the initial incubation time due to the higher
sulfate concentrations, as in low COD/sulfate conditions SRB always govern the organic matter
consumption (Jing et al., 2013). Methane production was enhanced only for COD/sulfate ratio
values approximately higher than 25-30 (Figures 2a-c), after which the deviation of electrons to
sulfate reduction apparently ceased.

Table 1. Kinetic parameters from the mathematical adjustment of the profiles obtained for methane production, SCOD
removal and sulfate removal.
COD/sulfate Methane production SCOD removal Sulfate removal
ratio
PM' a
RM' b
λ c
R2
CO SCOD d
CR SCOD d
k1SCOD' e
R2
COSO4 d k1SO4' e R2
12.0 (1st. 100.7 9.56 2.35 0.98 21644 14891 77.23 0.97 1854 38.01 0.98
fragment)
(Figure 1a) (Figure 1d) (Figure 1g)
12.0 (2nd. 538.1 6.51 9.44 0.99 12712 2932 23.87 0.97
fragment)
(Figure 1a) (Figure 1d)
10.0 (1st. 89.4 6.61 1.93 0.98 21639 13806 105.83 0.99 2118 36.46 0.97
fragment)
(Figure 1b) (Figure 1e) (Figure 1h)
10.0 (2nd. 375.6 7.97 23.02 0.98 14016 3183 31.88 0.95
fragment)
(Figure 1b) (Figure 1e)
7.5 (1st. 91.0 4.17 1.34 0.97 21584 12685 61.49 0.91 2649 29.00 0.92
fragment)
(Figure 1c) (Figure 1f) (Figure 1i)
7.5 (2nd. 261.9 9.39 43.7 0.99 12595 4103 28.24 0.99
fragment)
(Figure 1c) (Figure 1f)
Notes: aL kg-1VS; bL kg-1VS day-1; cday; dmg L-1; ekg-1VS day-1.

The methane production potential estimated for each condition (Table 1) also reflected the above
described behavior, indicating a value approximately 50% lower for the COD/sulfate ratio of 7.5
compared to the ratio of 12.0. Such discrepancy may also result from inhibitory effects of sulfide,
since higher concentrations were measured at the end of the incubation time for the reactors
submitted to lower COD/sulfate ratios (Table 2). pH values measured at the end of the experimental
run indicate the predominance of dissociate sulfide in the flasks (Table 2), regardless of the
COD/sulfate ratio. The literature reports a wide range of concentrations at which inhibitory effects
of sulfide are observed in anaerobic systems (50-400 and 100-800 mg L-1 respectively for
undissociated and dissociated sulfide, Chen et al., 2008), so that the definition of toxic levels
depends on intrinsic characteristics of each reactional system (Camiloti et al., 2014). However, the
patterns observed for the conversion of acetic acid also suggest the inhibition of the MA in the
reactors fed with higher concentrations of sulfate, especially for the COD/sulfate ratio of 7.5, as
further discussed.

Table 2. Overall performance in terms of pH, PA and sulfide concentration within the reactors.
COD/sulfate ratio pH PA TDS Ionized sulfide
(mgNaHCO3 L-1) (mg L-1) (mg L-1)
12.0 8.3 ± 0.0 7893 ± 55 114 ± 17 111 ± 16
10.0 8.4 ± 0.0 8391 ± 141 135 ± 2 132 ± 2
7.5 8.4 ± 0.0 8275 ± 212 194 ± 27 190 ± 27

Although lower methane production potential values were associated with higher sulfate
concentrations (Table 1), higher methane production rates were observed in the reactors initially fed
with lower COD/sulfate ratios, focusing on the periods when the deviation of the electron flow to
sulfidogenesis decreased to negligible patterns (second fragments, Figures 1a-c and Table 1). For
the COD/sulfate ratios of 10.0 and 7.5, RM’ reached respectively values about 20 and 40% higher
(second fragments, Figures 1b-c and Table 1) compared to the ratio of 12.0. This behavior may be
related to the oxidation of the organic matter to intermediate products, mainly acetate, by
incomplete SRB (Silva et al., 2011). The accumulation of acetic acid promoted by such bacteria
most likely enhanced the methanogenic activity by aceticlastic MA (A-MA) at sulfate shortage
conditions for the COD/sulfate ratios of 10.0 and 7.5 (second fragments, Figures 1b-c). The higher
partial alkalinity values observed for such conditions (Table 2) also promoted more suitable
conditions for the methane production via acetate conversion, partially offsetting eventual inhibitory
effects of sulfide. According to Godoi et al. (2015), some organic matter oxidation reactions
mediated by SRB generate bicarbonate alkalinity, which corroborate the patterns observed for PA.

Organic matter degradation and sulfate removal


Organic matter degradation temporal profiles for both experimental data and fitted models are
depicted in Figures 1d-f, whereas the parameters mathematically obtained are also summarized in
Table 1. First-order kinetic constants obtained for the COD removal at higher sulfate concentrations
(first fragments, Figures 1d-f and Table 1) reached values 2- to 3-fold higher compared to the
periods when the electron flow was not diverted from methanogenesis to sulfidogenesis (second
fragments, Figures 1d-f and Table 1). Such pattern results from discrepancies regarding kinetic and
substrate affinity characteristics between SRB and MA.

Aceticlastic sulfidogenesis (Reaction 1, Saady, 2013) probably governed the marked primary
decreasing pattern observed for COD removal in all flasks, also considering kinetic and
thermodynamic advantages of aceticlastic SRB (A-SRB) over aceticlastic MA (A-MA) (Reaction 2)
(Lens et al., 1998). The discrepancy between the first and second fragments in the COD removal
profiles is more evident for the reactors fed with higher sulfate concentrations (Figures 1e-f).
However, it should be mentioned that despite the decreasing concentrations of acetic acid observed
at the initial period of the incubation (Figure 3), the mechanisms that govern the competition
between A-SRB and A-MA are not yet fully understood (Lens et al., 1998; Chen et al., 2008).
Figure 1. Temporal profiles and mathematical fitting obtained for methane production -COD/sulfate = 12.0 (-○-, a), 10.0 (-□-,b) and 7.5 (-Δ-, c); SCOD removal -
COD/sulfate = 12.0 (-○-, d), 10.0 (-□-, e) and 7.5 (-Δ-, f) and sulfate removal -COD/sulfate = 12.0 (-○-, g), 10.0 (-□-, h) and 7.5 (-Δ-, i). Note: aRemoving sulfide interference.
Figure 2. COD/sulfate temporal variation (X) – COD/sulfate = 12.0 (a), 10.0 (b) and 7.5 (c), and electron flow for
sulfidogenesis and methanogenesis – COD/sulfate = 12.0 (d), 10.0 (e) and 7.5 (f). Note: aRemoving sulfide interference.

SO42- + CH3COOH → 2CO2 + S2- + 2H2O (1)


ΔGº’ = -47.6 kJ mol-1

CH3COOH → CH4 + H2O (2)


ΔGº’ = -31.0 kJ mol-1

At the beginning of the incubation period, the release of hydrogen (H2) by fermentative bacteria
during the acidification of the readily available organic matter from vinasse probably also
stimulated sulfate reduction by hydrogenotrophic SRB (H-SRB). In addition to presenting a lower
hydrogen threshold concentration, i.e. higher affinity to H2, (Chen et al., 2008), the free energy
obtained by H-SRB from H2 (Reaction 3) is approximately 13 and 46% higher compared to
hydrogenotrophic MA (H-MA) (Reaction 4) and homoacetogenic bacteria (Reaction 5),
respectively (Silva et al., 2011; Saady, 2013). The values presented herein support the results
previously discussed for methane production and further presented for acetate consumption, so that
the low patterns regarding the cumulative methane production (first fragments, Figures 1a-c)
suggest that SRB competed with MA by both acetate and H2.

SO42- + 4H2 + H+ → HS- + 4H2O (3)


ΔGº’ = -152.2 kJ mol-1

4H2 + CO2 → CH4 + 2H2O (4)


ΔGº’ = -135.0 kJ mol-1

4H2 + 2CO2 → CH3COOH + 2H2O (5)


ΔGº’ = -104.0 kJ mol-1

The establishment of sulfate shortage conditions (COD/sulfate approximately higher than 25-30,
Figures 2a-c) and the accumulation of acetate from the metabolism of incomplete SRB (Figure 3)
probably enhanced the organic matter removal process to the aceticlastic methanogenic via in all
flasks, as previously discussed. Therefore, the comparison between the first-order kinetic
parameters obtained for the first and second fragments (Table 1), regardless of the COD/sulfate
ratio, should be carefully considered, based on the most likely activity of different microbial
populations and the utilization of different electron donors, i.e., hydrogen and acetate, by each
group. Nevertheless, Camiloti et al. (2014) reported a similar finding by studying the simultaneous
sulfate and organic matter removal from complex synthetic wastewater in a continuous fixed-bed
reactor. At a COD/sulfate ratio of 6.1, the authors observed a higher kinetic constant for COD
removal via sulfidogenic conditions compared to the methanogenic process. Such results indicate
that the consumption of fermentation products may be more favorable to incompletely oxidizing
SRB than to MA, depending on the availability of sulfate in the reactor.

First-order kinetic constants associated with the organic matter removal at COD/sulfate ratios of
10.0 (31.88 kg-1VS d-1) and 7.5 (28.24 kg-1VS d-1), after sulfate shortage conditions (second
fragments, Figures 1e-f and Table 1), were slightly higher compared to the control flasks, i.e.,
COD/sulfate = 12.0 (23.87 kg-1VS d-1, second fragment, Figure 1a and Table 1). This finding
corroborates the patterns observed for the kinetic parameters from methane production, so that the
higher PA values and acetate accumulation most likely favored the aceticlastic methanogenic
activity, as previously discussed.

The kinetic constants obtained for the sulfate removal are also summarized in Table 1, whereas the
temporal profiles of sulfate concentration decay considering both experimental data and fitted
models are depicted in Figures 1g-i. k1SO4’ values slightly decreased as the initial sulfate
concentration increased, i.e., lower COD/sulfate ratios, which may result from several factors (Chen
et al., 2008; Camiloti et al., 2014): reduced availability of electron donors to sulfidogenesis,
increasing concentrations of sodium due to the addition of Na2SO4 or sulfide inhibition. The
relatively low sulfate concentrations applied in this study compared to the moderate-to-high
availability of organic compounds suggest that adverse effects from electron donors shortage may
be considered negligible, considering the stoichiometric COD/sulfate ratio of 0.67, at which organic
matter may be completely oxidized by SRB (Lens et al., 1998). The complex compositional nature
of vinasse also limits the identification of inhibitory effects from specific compounds in this study.
Figure 3. Temporal profiles for VFA concentration – COD/sulfate = 12.0 (a), 10.0 (b) and 7.5 (c), and proportion –
COD/sulfate = 12.0 (d), 10.0 (e) and 7.5 (f).

VFA production and consumption


Figure 3 depicts the temporal profiles for VFA – acetic (AcH), propionic (PrH) and butyric (BuH)
acids – concentration and proportion at each COD/sulfate ratio studied. AcH concentration decay
(Figure 3a) presented a similar pattern compared to the temporal profile of SCOD removal (Figure
1d) at a COD/sulfate ratio of 12.0. In this case, a marked punctual decrease may be observed in
AcH concentration closely to 20th day of incubation, followed by a subsequent increase and then a
continuous decay up to concentrations lower than 200 mg L-1 (Figure 3a). The above described
punctual decrease most likely resulted from the consumption of the AcH primarily present in the
raw vinasse, whereas the further increase in the concentration of AcH resulted from the metabolism
of incomplete SRB. PrH concentrations presented a slight increase at the beginning of the
incubation period, which comprises a common characteristic of thermophilic anaerobic systems
(Speece et al., 2006). According to Wiegant et al. (1985) the degradation of PrH, i.e., conversion to
AcH (Reaction 6), may be considered the rate-limiting step in thermophilic reactors, requiring more
energy compared to the conversion of BuH (Reaction 7) (Saady, 2013). Such more favorable
thermodynamic characteristics corroborate the marked decrease observed for the BuH concentration
in the flasks (< 100 mg L-1), regardless of the COD/sulfate ratio. Yet regarding the dynamics of PrH
consumption at a COD/sulfate ratio of 12.0, a continuous decay was observed after AcH levels
reached values below 200 mg L-1 (Figure 3a). Such pattern suggests that PrH was converted to AcH
and then rapidly consumed by A-MA, based on the terminal increase observed for the methane
production (Figure 1a). Negligible variations were observed in both VFA concentration (Figure 3a)
and proportion (Figure 3d) approximately after the 60th day of incubation (COD/sulfate = 12.0), as
the low levels of each organic acid kinetically limited their conversion mediated by acetogenic
bacteria or A-MA.

CH3CH2COOH + 2H2O → CH3COOH + 3H2 + CO2 (6)


ΔGº’ = +76.2 kJ mol-1

CH3(CH2)2COOH + 2H2O → 2CH3COOH + 2H2 (7)


ΔGº’ = +48.4 kJ mol-1

A similar pattern was observed for the dynamics of AcH and PrH production and consumption in
the flasks with higher initial sulfate levels (COD/sulfate of 10.0 and 7.5). AcH concentrations
primarily presented a decay in both cases, followed by periods of slight (COD/sulfate =10.0, Figure
3b) to marked (COD/sulfate = 7.5, Figure 3c) accumulation after sulfate shortage conditions were
established (COD/sulfate higher than 25-30, Figure 2b-c). Simultaneously to the buildup of AcH a
marked decrease in the concentrations of PrH was observed in both cases, i.e. approximately 45th
and 60th days of incubation respectively for the COD/sulfate concentrations of 10.0 (Figure 3b) and
7.5 (Figure 3c). Such enhanced AcH production most likely resulted from the activity of incomplete
oxidizing SRB, which are also able to oxidize PrH to AcH in the absence of electron-acceptors, i.e.,
sulfur compounds (Vilela et al., 2014). SRB present higher affinity for propionate; moreover, PrH
oxidation reactions mediated by SRB are thermodynamically more favorable compared to the
metabolism of syntrophic acetogens (Chen et al., 2008; Barrera et al., 2014). VFA profiles obtained
at a COD/sulfate ratio of 12.0 corroborate such hypothesis: as previously discussed, PrH
degradation was enhanced only after AcH markedly decreased at a COD/sulfate ratio of 12.0
(Figure 3a), whereas for the COD/sulfate ratios of 10.0 (Figure 3b) and mainly 7.5 (Figure 3c) PrH
degradation took place prior to consumption of AcH up to the terminal concentrations.

Finally, the relatively high residual AcH concentrations observed at the lowest COD/sulfate ratio
(7.5, Figure 3c) suggest that A-MA were submitted to inhibiting conditions, which corroborates the
lower performances obtained for both organic matter removal and methane production. Barrera et
al. (2014) reported the inhibition of both MA and SRB by sulfide at COD/sulfate ratios ranging
from 10.0 to 5.0 in an upflow anaerobic sludge blanket reactor continuously fed with sugarcane
vinasse supplemented with Na2SO4. In this case, the authors observed the accumulation of AcH to
levels higher than 3000 mg L-1, with undissociated and dissociated sulfide concentrations
respectively higher than 750 and 160 mg L-1. Based on the pH values measured in the flasks at the
end of the experimental run (≥ 8.3, Table 2), sulfide was predominately in the dissociated form in
the liquid phase, with values ranging from 111 (COD/sulfate = 12.0) to 190 mg L-1 (COD/sulfate =
7.5). Although both concentrations are within the range of toxic levels defined for dissociated
sulfide in anaerobic environments (100-800 mg L-1, Chen et al., 2008), sulfide concentration at a
COD/sulfate ratio of 7.5 was approximately 70% higher compared to the one at a ratio of 12.0.
Since lower dissociated sulfide concentrations led to inhibitory effects over the granular sludge in
Barrera et al. (2014), the hypothesis of inhibition by sulfide in the flasks submitted to higher sulfate
concentrations is acceptable, because the suspended cell growth and the mixing conditions
theoretically enhanced the contact of the cells with sulfide. Mass transfer resistance in immobilized-
cell systems tends to enhance the tolerance of anaerobic microbial populations to sulfide (Lens et
al., 1998); therefore, future studies under such conditions should also be useful to better understand
the toxicity character of sulfide within the anaerobic digestion of sugarcane vinasse.

CONCLUSIONS
The results obtained in this study indicated that the higher the sulfate concentration, the lower the
methane production from vinasse, which suggests that the methanogenic activity was negatively
impacted by sulfidogenesis. At a COD/sulfate ratio of 7.5 methane production potential was
approximately 35% lower compared to a ratio of 12.0, considering the deviation of about 13.6% of
the electron flow to sulfidogenesis. Organic matter consumption was also hindered under such
conditions, most likely as result from inhibitory effects of the sulfide (approximately 200 mg L-1)
accumulated throughout the incubation period; the buildup of acetic acid corroborates such
hypothesis of inhibition by sulfide at a COD/sulfate ratio of 7.5. VFA profiles also suggest that
sulfate reduction was carried out by incompletely oxidizing SRB in the flasks, so that the
accumulation of acetate and the higher levels of bicarbonate alkalinity enhanced the COD
conversion rates when methanogenesis governed the electron flow at lower COD/sulfate ratios.
COD and sulfate removal profiles indicated that deviation of electrons to sulfidogenesis was
considered negligible at COD/sulfate ratios approximately higher than 25, which was evidenced by
the exponential increase in methane production. At a COD/sulfate ratio of 12.0 the electron flow
diverted to sulfidogenesis was lower than 7.6%, which slightly impaired methane production. In
short, the results obtained herein indicate that AD systems applied to sugarcane vinasse tend to be
negatively impacted throughout the sugarcane harvest, based on the accumulation of sulfate in
vinasse due to the continuous use of sulfuric acid in biorefineries. Consequently, the operation of
such systems may be conducted from different perspectives, considering approaches without fixing
design and operating parameters or designing the AD plant for the worst scenario, i.e., when higher
sulfate levels are observed in vinasse (end of the harvest).

ACKNOWLEDGEMENTS
The authors are grateful to the São Paulo Research Foundation (Fapesp), Grant numbers
2009/15984-0, 2012/15606-9 and 2014/19091-9, for supporting the development of this study, and
to Mr. Claudio Frazão Araújo and Ms. Michele Fernanda Pires, from Usina São Martinho, for
providing the sugarcane vinasse samples used herein.

References
Adorno, M.A.T., Hirasawa, J.S. and Varesche, M.B.A. (2014) Development and validation of two methods to quantify volatile acids
(C2-C6) by GC/FID: headspace (automatic and manual) and liquid-liquid extraction (LLE). AJAC 5, 406–414.
Barrera, E.L., Spanjers, H., Romero, O., Rosa, E. and Dewulf, J. (2014) Characterization of the sulfate reduction process in the
anaerobic digestion of very high strength and sulfate rich vinasse. Chem. Eng. J. 248, 383–393.
Camiloti, P.R., Mockaitis, G., Rodrigues, J.A.D., Damianovic, M.H.R.Z., Foresti, E. and Zaiat, M. (2014) Innovative anaerobic
bioreactor with fixed-structured bed (ABFSB) for simultaneous sulfate reduction and organic matter removal. J. Chem. Technol.
Biotechnol. 89, 1044–1050.
Chen, Y., Cheng, J.J. and Creamer, K.S. (2008) Inhibition of anaerobic digestion process: a review. Bioresour. Technol. 99, 4044–
4064.
Costa, R.B. and Foresti, E. (2014) Strategies for the enrichment of methanotrophic denitrification community. In: Proceedings of the
XI Latin American Workshop and Symposium on Anaerobic Digestion, La Havana, Cuba, pp. 804–811.
Costa, F.J.C.B., Rocha, B.B.M., Viana, C.E. and Toledo, A.C. (1986) Utilization of vinasse effluents from an anaerobic reactor. Wat.
Sci. Tech. 18(12), 135–141.
Ferreira, L.F.R., Aguiar, M.M., Messias, T.G., Pompeu, G.B., Lopez, A.M.Q., Silva, D.P. and Monteiro, R.T. (2011) Evaluation of
sugac-cane vinasse treated with Pleurotus sajor-caju utilizing aquatic organisms as toxicological indicators. Ecotox. Environ.
Safe. 74, 132–137.
Fuess, L.T. and Garcia, M.L. (2015) Bioenergy from vinasse anaerobic digestion to enhance the energy balance ratio of ethanol
production. J. Environ. Manage. 162, 102–114.
Godoi, L.A.G., Damianovic, M.H.R. and Foresti, E. (2015) Sulfidogenesis interference on methane production from carbohydrate-
rich wastewater. Wat. Sci. Tech. 72, 1644–1652.
Jing, Z., Hu, Y., Niu, Q., Liu, Y., Li, Y.Y. and Wang, X.C. (2013) UASB performance and electron competition between methane-
producing archaea and sulfate-reducing bacteria in treating sulfate-rich wastewater containing ethanol and acetate. Bioresour.
Technol. 137, 349–357.
Kumar, G.S., Gupta, S.K. and Singh, G. (2007) Biodegration of distillery spent wash in anaerobic hybrid reactor. Water Res. 41,
721–730.
Lebrero, R., Toledo-Cervantes, A., Muñoz, R., Del Nery, V. and Foresti, E. (2015) Biogas upgrading from vinasse digesters: a
comparison between an anoxic biotrickling filter and an algal-bacterial photobioreactor. J. Chem. Technol. Biotechnol., doi:
10.1002/jctb.4843.
Lens, P.N.L., Visser, A., Janssen, A.J.H., Hulshoff Pol, L.W. and Lettinga, G. (1998) Biotechnological treatment of sulfate-rich
wastewaters. Crit. Rev. Env. Sci. Tec. 28(1), 41–88.
Ripley, L.E., Boyle, W.C. and Converse, J.C. (1986) Improved alkalimetric monitoring for anaerobic digestion of high-strength
wastes. J. Water Pollut. Control Fed. 58, 406–411.
Saddy, N.M.C. (2013) Homoacetogenesis during hydrogen production by mixed cultures dark fermentation. Int. J. Hydrogen Energ.
38, 13172–13191.
Silva, A.J., Domingues, M.R., Hirasawa, J.S., Vareche, M.B., Foresti, E. and Zaiat, M. (2011) Kinetic modeling and microbial
assessment by fluorescent in situ hybridization in anaerobic sequencing batch biofilm reactors treating sulfate-rich wastewater.
Braz. J. Chem. Eng. 28(2), 209–219.
Speece, R.E., Boonyakitsombut, S., Kim, M., Azbar, N. and Ursillo, P. (2006) Overview of anaerobic treatment: thermophilic and
propionate implications. Water Environ. Res. 78, 460–473.
Standard Methods for the Examination of Water and Wastewater 2005 21st edn, American Public Health Association/American
Water Works Association/Water Environment Federation, Washington DC, USA.
Vela, F.J., Zaiat, M. and Foresti, E. (2002) Influence of the COD to sulphate ratio on the anaerobic organic matter degradation
kinetics. Wat. SA 28, 213–216.
Vilela, R.S., Damianovic, M.H.R.Z. and Foresti, E. (2014) Removing organic matter from sulfate-rich wastewater via sulfidogenic
and methanogenic pathways. Wat. Sci. Tech. 69, 1669–1675.
Wiegant, W.M., Claassen, J.A. and Lettinga, G. (1985) Thermophilic anaerobic digestion of high strength wastewaters. Biotechnol.
Bioeng. 27, 1374–1381.
Wilkie, A.C., Riedesel, K.J. and Owens, J.M. (2000) Vinasse characterization and anaerobic treatment of ethanol vinasse from
conventional and cellulosic feedstocks. Biomass Bioenerg. 19, 63–102.
Zwietering, M.H., Jongenburger, I., Rombouts, F.M. and van’t Riet, K. (1990) Modeling of the bacterial growth curve. Appl.
Environ. Microb. 56(6), 1875–1881.

You might also like