2019 Scienceline Publication

Journal of Civil Engineering and Urbanism

Volume 9, Issue 4: 36-42; July 25, 2019

ISSN-2252-0430

DOI: https://dx.doi.org/10.29252/scil.2019.jceu5

Predicting Bioenergy Potential from Vinasse Digestion: The

VUMP Model (Vinasse Utilization for Methane Production)

Lucina Marcia Kuusisto¹, Melanie L. Sattler²*, Victoria C.P. Chen³

¹Department of Biological and Environmental Sciences, Texas A&M University - Commerce, 2200 Commerce St., Commerce, TX 75429 USA

²Department of Civil Engineering, University of Texas at Arlington, Box 19308, Arlington, TX 76019 USA

³Department of Industrial and Manufacturing Systems Engineering, University of Texas at Arlington, Box 19308, Arlington, TX 76019 USA

^Corresponding author’s Email: sattler@uta.edu

ABSTRACT

Global ethanol production generates 900 to 2000 billion liters per year of a high-strength liquid waste called

vinasse. Vinasse represents a substantial renewable energy resource, through anaerobic digestion to produce biogas.

Although a variety of previous studies have measured biogas generation from anaerobic treatment of vinasse, no

previous study has developed a general model to predict biogas production from anaerobic digestion of vinasse of

any composition at a range of temperatures (mesophilic). The aim of this research was thus to build a first-order

model VUMP (Vinasse Utilization for Methane Production) to predict methane generation from anaerobic treatment

of vinasse from ethanol production, based on readily available inputs of initial vinasse composition and treatment

temperature. Lab-scale anaerobic digesters were filled with 4 synthetic vinasse mixtures with differing initial values

of chemical oxygen demand, nitrogen, potassium, phosphorous, and sulfur, operated at 3 temperatures each, for a

total of 12 reactors. Based on data collected, a multiple linear regression equation (R²= 0.80) was developed to

predict first-order methane generation rate constant k. The selected best-fit model for k varied positively as

functions of temperature, initial chemical oxygen demand, and the product of nitrogen and phosphorous.

Preliminary validation indicated that the model predicted methane generation from commercial vinasse within 20%.

Keywords: Anaerobic Digestion, Biogas, Ethanol, Methane

INTRODUCTION

2000; Baez-Smith, 2006; España-Gamboa et al., 2011;

Moraes et al., 2015 ).

Ethanol presents many advantages as a biofuel. However,

the production of ethanol from biomass generates a high-

strength liquid waste called vinasse. Per L of ethanol, 9-20

L of vinasse can be generated, depending on the feedstock

(corn, sugar crops, starch crops, dairy products, or

cellulosic materials) (Wilkie et al., 2000; Parnaudeau et

al., 2008; España-Gamboa et al., 2011 ). Globally, almost

100 billion L of ethanol were produced in 2015. This

generated 900 to 2000 billion L of vinasse (US

Department of Energy, 2016 ).

Traditionally, Brazil, Australia, and other countries

have disposed of vinasse by applying it as a fertilizer on

agricultural land (Korndorfer and Anderson, 1997;

Moraes et al., 2015 ). This can produce short-term benefits,

because the vinasse contains nutrients like potassium,

magnesium, and calcium which are needed for crops like

sugarcane (España-Gamboa et al., 2011 ). However, over

the long term, such disposal can cause severe deterioration

of soil, surface water and ground water, due to vinasse’s

high biochemical oxygen demand (BOD) and chemical

oxygen demand (COD) (20-60 g/L and 50-150 g/L,

respectively); low pH (3.5-5); and high concentrations of

solids (30-70 g/L), nitrogen (300-800 mg/L), phosphorous

(100-500 mg/L), and potassium (2-3 g/L) (Wilkie et al.,

Satyawali and Balakrishnan (2008) reviewed the

existing status and advances in treatment methods for

vinasse, and found that anaerobic treatment was the most

attractive primary treatment due to 80% BOD removal.

Studies have shown that anaerobic treatment reduces

vinasse COD by 67–98% (Vijayaraghavan and

Ramanujam, 2000 ). Anaerobic treatment not only reduces

organic pollutants but also produces stabilized residuals

that can be used as fertilizer without creating water

pollution problems. Finally, anaerobic treatment produces

renewable energy in the form of biogas (Mota et al.,

2015). The general process of anaerobic degradation for

any organic substrate can be represented as (Deublein and

Steinhauser, 2008 ):

C_cH_hO_oN_nS_s+ yH₂O  xCH₄+ (c-x)CO₂+ nNH₃+ sH₂S (1)

where

x = 1/8 * (4c+h -20 –3n–2s) and y = ¼ * (4c–h–20+3n+3s).

A variety of previous studies have measured

methane generation from anaerobic treatment of vinasse

(dos Reis et al., 2015; de Barros et al., 2017; Fu et al.,

2017). However, only limited number of studies have

attempted to predict or model methane generation. A few

of these previous studies estimated kinetic parameters for

methane production from a single vinasse digested at a

To cite this paper: Kuusisto LM, Sattler ML, Chen VCP (2019). Predicting Bioenergy Potential from Vinasse Digestion: The VUMP Model (Vinasse Utilization for Methane

Production). J. Civil Eng. Urban., 9 (4): 36-42. www.ojceu.ir

single temperature (Zieminski and Kowalska-Wentel,

2015; Albanez et al., 2016 ). Vinasse composition,

however, varies based on the feedstock utilized to produce

the ethanol, as well as the particular production process

(Wilkie et al., 2000 ). Several previous studies modeled

how methane production varied as a function of one

parameter: vinasse COD/N ratio (Syaichurrozi et al.,

2013), COD/sulfate ratio for sugarcane vinasse (Kiyuna et

al., 2017 ), or ratio of sugarcane press mud to vinasse

(González et al., 2017 ). Baez-Smith (2006) developed a

method to estimate methane generation from vinasse by

assuming 90% BOD removal.

No previous study, however, has developed a more

general model to predict methane generation from

anaerobic digestion of vinasse of any composition at a

range of temperatures. It was hypothesized that the initial

vinasse COD concentration would affect the methane

production rate, since non-zero-order reaction rates are

functions of reactant concentrations. Glucose, which is

converted by microorganisms to methane, provides the

COD. Other initial constituents could potentially affect

microbial health and growth and thus impact reaction

rates. It was also hypothesized that digestion temperature

would impact reaction rates, since microbial-facilitated

reactions typically increase with temperature, up to an

optimal range.

easy to obtain. A more complex model may provide more

accurate estimates, but at the cost of additional data

inputs. First-order models are used with great success to

describe many processes of interest in environmental

engineering; they often represent a reasonable balance

between simplicity and accuracy (Cooper, 2015 ). Even

though they do not account for all of the multiple steps

that occur in anaerobic degradation processes, first-order

models are widely used. For example, the US

Environmental Protection Agency’s widely used Landfill

Gas Emissions Model (LandGEM) is a first-order model.

Accordingly, an objective of this research was to develop

a first-order model to predict methane generation from

anaerobic treatment of vinasse.

MATERIAL AND METHODS

Vinasse compositions tested

In this research, synthetic compositions were used to

enable us to vary the range of constituent values tested,

and the ratios of constituent values, to aid in building a

model applicable to

wider range of vinasse

compositions. An experimental design (strength

orthogonal array) was developed (Bose and Bush, 1952;

Chen, 2001 ), covering the range of vinasse constituents

described in the literature (Wilkie et al., 2000 ), but

limiting the constituents to levels not toxic to

methanogens. Preliminary tests were conducted using the

experimental design; however, only 4 of 18 batches

produced significant methane. The other vinasse

compositions likely contained constituent quantities and

ratios unsuitable for microbial growth. The four vinasse

compositions used for full testing and model building are

shown in Table 1. For comparison, vinasse constituent

values reported in the literature are shown in the second

row from the bottom.

The overall goal of this research was to test these

hypotheses, and to build a model VUMP (Vinasse

Utilization for Methane Production) to predict methane

generation for anaerobic treatment of vinasse from ethanol

production, based on readily available inputs of initial

vinasse composition and treatment temperature. Specific

research objectives were:

1. To develop and operate laboratory scale anaerobic

reactors to study the effect of vinasse composition and

temperature on methane generation over time.

2. Using the laboratory data, to develop multiple-

linear regression equations for predicting first-order rate

constants for methane generation in terms of initial

vinasse composition and temperature.

Temperatures tested

To study the effect of temperature, tests were

conducted at 3 temperatures (30, 35 and 40°C), spanning

the range of mesophilic methanogens. Each of the 4

compositions was operated at 3 mesophilic temperatures

(30, 35, and 40°C), for a total of 12 runs.

Anaerobic systems can be designed for temperatures

appropriate for mesophilic bacteria (30-40°C) or

thermophilic bacteria (50-60°C). Higher temperatures

increase microbial activity, with activity roughly doubling

for every 10°C increase within the optimal range (Khanal,

2008). Thermophilic systems thus produce methane 25-

50% faster, depending on the substrate (Henze and

Harremoes, 1983 ). The thermophilic range also

demonstrates improved pathogen destruction. However,

start-up for thermophilic systems is slower, and systems

are more susceptible to changes in loading variations,

substrate, or toxicity (Khanal, 2008 ). Hence, for this

research, mesophilic temperatures were chosen, since

these systems operate with more stability.

Batch reactor set-up

Experiments were conducted in 3-L glass reactor

flasks. Each reactor was connected to an air-tight gas-

collection bag (22-L Cali-5-Bond™ Bag, Calibrated

Instruments, Inc.), as shown in Figure 1. Before filling, all

reactors were sealed with silicon sealant and leak-

checked. Reactors were filled with vinasse of different

compositions, according to Table 1. The synthetic vinasse

mixtures were prepared using appropriate concentrations

of glucose, ammonia, phosphoric acid, potassium

hydroxide, and calcium sulfate as sources of chemical

oxygen demand (COD), nitrogen (N), phosphorous (P),

To facilitate their use, the models should be based

on a limited number of input parameters that are fairly

To cite this paper: Kuusisto LM, Sattler ML, Chen VCP (2019). Predicting Bioenergy Potential from Vinasse Digestion: The VUMP Model (Vinasse Utilization for Methane

Production). J. Civil Eng. Urban., 9 (4): 36-42. www.ojceu.ir

potassium (K), and sulfur (S), respectively. The pH was

adjusted to within between 7.0 and 8.4 using hydrochloric

acid or sodium hydroxide as appropriate, and then sodium

bicarbonate was added to buffer the pH. Anaerobic-

digested sewage sludge collected from the City of Fort

Worth Village Creek Wastewater Treatment Plant was

added to inoculate each bioreactor with an initial supply

of microbes. Enough sludge was added to comprise 10-

15% of the total solution volume, according to Espinoza-

Escalante et al . (2008). Revised Anaerobic Mineral

Medium (RAMM) (Shelton and Tiedje, 1984 , modified

for use with vinasse by omitting constituents with N, P, or

K) was added to each batch to ensure that microbes had

sufficient minor nutrients.

Figure 1. Experimental set-up

Table 1. Vinasse compositions tested

Composition

COD, mg/L

N (as NH₃), mg/L

P (as PO₄)

S (as SO₄)

1 - Low COD

2,600

75,000

1200

580

2 – Medium COD, Low P, K

3 – High COD

147,000

550

580

4 – Medium COD, High P, K

Range of values tested

Literature values ^a,b

75,000

1200

1742

2,600 to 147,000

14,000 to 147,000

Very low to high

60 to 1200

56 to 13,760 (total N)

Low to moderate

7 to 90

39 to 1742

38.5 to 14,500

Low to moderate

34 to 580

34-9500

Low to moderate

0.68 to 1,990 (total P)

Low to moderate

Values tested compared to literature

a,b

España-Gamboa et al. (2012); Wilkie et al. (2000).

Where:

Reactor operation and monitoring

The reactors were operated at the 3 different

mesophilic temperatures (30, 35, 40 °C) via placement in

constant-temperature rooms. Magnetic stirrers were used

to provide continuous mixing. The reactors were operated

until methane generation ceased (5-10 days).

V= cumulative volume of methane per liter of

vinasse (mL/L),

L_o= ultimate methane potential (mL/L),

k = first-order methane generation rate constant

(day¹),

During the initial stages of hydrolysis and

acidogenesis, sodium hydroxide was added every 3 hours

so that the pH did not fall below 5, which would have

caused methanogens to die. Later, the pH was adjusted

less frequently, as needed. The biogas volume was

measured daily by pumping the gas out of the collection

bag through a standard air-grab sampler (SKC Air-check

pump, model 224‐44XR), which pumped the biogas at 1.0

L/min, and was connected to a calibrator (Bios Defender

510M). During the gas pumping period, the time needed

to empty the gas bag was recorded. A LANDTEC-GEM

2000 PLUS with infrared gas analyzer (±3% accuracy)

was used to measure the concentration of methane in

percent volume. LANDTEC measurements of methane

have previously been compared to those from a gas

chromatograph, and found to be within 7% of the GC

readings (Karanjekar, 2012 ).

t = time (days).

Rearranging Eq. 1 and taking the natural log of both

sides gives:

ln(1-V/L_o) = -kt

(2)

If ln(1-V/L_o) is plotted vs. time, the negative value

of the slope gives k. Linear regression to determine k

values for each experiment was performed using MS

Excel software. L_owas estimated from the horizontal

asymptote of the plots of ln(1-V/L_o) vs. time. When the

plot did not clearly reach an asymptote, the value of L_o

was chosen which gave the largest R²value for a

regression line fit to ln(1-V/L_o) vs. time.

Based on the 12 k values (one for each bioreactor

run), a comprehensive multiple linear regression (MLR)

equation was developed using SAS software. Six predictor

variables (temperature and the five waste components

COD, N, P, K, and S) were used to estimate k as the

response variable. The following steps were followed for

developing each MLR equation: reviewing raw data plots

and correlation analyses; developing a preliminary MLR

model and checking model assumptions; conducting

remedial actions, such as transformations, until the model

assumptions for regression analysis were satisfied;

Methane model development

Assuming methane generation is first-order,

cumulative methane volume can be estimated using the

following equation:

V = L_o(1-e^-kt)

(1)

To cite this paper: Kuusisto LM, Sattler ML, Chen VCP (2019). Predicting Bioenergy Potential from Vinasse Digestion: The VUMP Model (Vinasse Utilization for Methane

Production). J. Civil Eng. Urban., 9 (4): 36-42. www.ojceu.ir

exploring possible interaction terms; searching for good

fit MLR models; selecting the best-fit MLR model.

Checks were performed for constant variance (modified

Levene test), normal distribution of residuals, outlier

influence (Bonferroni test), multi-collinearity (variance

inflation values). Transformation of the data was not

necessary since normality and constant variance tests

proved to be satisfactory.

The best model was selected using the backward

elimination method, best subsets method and stepwise

regression method, such that all parameters were

significant at α = 0.1. The best-fit model was selected

based on the R²value, adjusted R²value, Mallows C_p, and

Akaike Information Criterion (AIC) or Schwarz Bayesian

Criterion (SBC). Parsimonious models with high R²and

adjusted R², low Mallows C_p, and low AIC or SBC were

used because they represent the overall goodness of fit

and avoid unnecessary predictor variables (Kutner and

Neter, 2004).

fastest for 40°C, intermediate for 35°C, and slowest for

30°C. For each composition (with the exception of

medium COD with low K and P, which had atypically low

methane generation for an unknown reason), the

maximum value of cumulative methane generation is

similar for the three temperatures. This is expected

because the temperature impacts only the rate of methane

generation. The maximum cumulative amount depends

only on initial COD (glucose), which is the same for a

given composition.

RESULTS AND DISCUSSION

Reactor data

Figure 2 compares cumulative methane generated

per liter of vinasse for the various compositions at 30°C,

35°C, 40°C. At each temperature, the composition with

the lowest initial COD generates the least methane, and

the composition with the highest initial COD generates the

greatest methane, as expected. At each temperature, the

composition with medium COD with high values of K and

P generates more methane than the composition with

medium COD with low values of K and P. According to

the regression model discussed later, higher amounts of P

are favorable for methane generation, presumably because

they favor microbial growth/metabolism.

Despite the fact that the initial high COD (147 g/L)

is almost double that of the medium COD compositions

(75 g/L), methane generation for the high COD

composition is not double that of the either medium COD

composition, for any of the temperatures (with the

exception of medium COD with low K and P, which had

low methane generation). The medium COD compositions

have higher N than the high COD composition. According

to the regression model discussed later, higher amounts of

N are favorable for methane generation, presumably

because they favor methane growth/metabolism.

a) 30°C

b) 35°C

c) 40°C

Methane generation for the composition with low

COD (2.6 g/L) is higher than would be expected,

considering that its COD value is only 3.5% that of the

medium value, and 1.8% that of the high value, and that

its N and P values are also low. This may be due to

increase in the initial COD due to the sludge addition for

microbial seeding.

Figure 3 compares cumulative methane generated

vs. time for each composition at the three temperatures.

As expected, for each composition, methane generation is

Figure 2. Cumulative methane generated from four

compositions at a) 30°C, b) 35°C, c) 40°C

To cite this paper: Kuusisto LM, Sattler ML, Chen VCP (2019). Predicting Bioenergy Potential from Vinasse Digestion: The VUMP Model (Vinasse Utilization for Methane

Production). J. Civil Eng. Urban., 9 (4): 36-42. www.ojceu.ir

The R²values for the other curve fits were above 0.75 for

all reactors.

30°C

35°C

40°C

Table 2. Rate constant k values for the anaerobic reactors

Temp.

(⁰C)

Composition

(/day)

N/A

0.67

0.52

0.61

0.62

0.70

0.74

0.88

0.56

0.57

0.79

Day

a) Composition with initial low COD

30°C

35°C

40°C

Day

Model development

b) Composition with initial medium COD, high K, P

Using the 10 k values shown in Table 2, an MLR

model was developed using a rigorous MLR procedure

described in detail by Kuusisto (2013). The selected best

model for k is shown below in Eq. 3 (R²= 0.87, adjusted

R²= 0.80).

30°C

35°C

40°C

k = - 4.96822 + 0.00243COD + 0.01757T + 0.05107(N x P) (3)

where:

k = methane generation first-order rate constant (day^-1);

COD = Chemical Oxygen Demand (g/L);

T = Temperature in the mesophilic range (K);

N = Nitrogen concentration (g/L);

P = Phosphorus concentration (g/L)

Day

All terms in the regression model were significant at

α = 0.1 level (p-values were less than significance level).

Real vinasse was obtained from White Energy

Ethanol Distillery to perform a limited validation of the

model. The vinasse solution had the following

composition: COD = 3.171 g/L; N = 0.0662 g/L; P =

0.2887 g/L. The VUMP model was used to calculate a

k_calculatedvalue of 0.452 per day. A sample of the vinasse

solution was then digested at 35°C, and methane volume

generated was measured over time. Using the methane

volume data and Eq. 2, k_actualwas determined to be

0.378/day. The model over-predicted the actual first-order

rate constant by 20%. This was likely because the

substrate used for model development was glucose, which

is likely easier for microorganisms to digest than many of

the compounds in real vinasse. Additional model

validation is recommended.

c) Composition with initial high COD

30°C

35°C

40°C

Day

d) Composition with initial medium COD, low K, P

Figure 3. Cumulative methane generated at three

temperatures for four vinasse compositions

Using the data for methane generation vs. time, rate

constant k values were determined for each reactor using

Eq. (2), as discussed above. Table 2 shows the k values for

the 12 reactor runs. k for Composition 1 at 30°C and 35°C

are shown as “N/A,” because they only weakly followed a

first-order trend (R²< 0.75), for reasons that are unclear.

Trends in the regression model for k

Eq. 3 shows that k increases with COD, temperature,

and the interaction between N and P. k was not a function

of K or S. The dependence on COD concentration was not

surprising, given that COD is an indirect indicator of

organic matter content, and organic carbon is converted to

To cite this paper: Kuusisto LM, Sattler ML, Chen VCP (2019). Predicting Bioenergy Potential from Vinasse Digestion: The VUMP Model (Vinasse Utilization for Methane

Production). J. Civil Eng. Urban., 9 (4): 36-42. www.ojceu.ir

methane in anaerobic digestion. The temperature

dependence is not surprising either, given that microbial

activity typically varies as a function of temperature

(Khanal, 2008 ).

The fact that the equation was not a function of S

was surprising. Kiyuna et al. (2017) found that higher

vinasse sulfate concentrations decreased methane

production. Barrera et al. (2014) mention the negative

impact of sulfide on methanogens; although S was

measured as sulfate in this study, a portion of sulfate was

likely converted to sulfide by sulfate-reducing bacteria,

since conditions were anaerobic.

The coefficient of the NxP term (0.05107) is larger

than the coefficient of the COD term (0.00243), indicating

that the methane generation rate will vary most strongly

with changes in initial N or P content. Syaichurrozi et al.

(2013) similarly found that methane production varied

with vinasse COD/N ratio, with the optimal ratio being

600:7.

DECLARATIONS

Acknowledgments

This research did not receive any specific grant from

funding agencies in the public, commercial, or not-for-

profit sectors.

Authors’ contributions

Lucina Marcia Kuusisto developed the experimental

design, conducted the experiments, and developed the

regression model. Melanie Sattler supervised Dr.

Kuusisto’s work and wrote this manuscript. Victoria Chen

helped with the experimental design and regression model

building. All authors read and approved the final

manuscript.

Competing interests

The authors declare that they have no competing

interests.

Consent to publish

Not applicable

Model limitations

In this study, a model for the first-order rate constant

for methane generation from vinasse was developed using

lab-scale data. The model represents a first step toward the

goal of being able to estimate methane generation from

anaerobic biological treatment of vinasse of any

composition at mesophilic temperatures, using a limited

number of predictor variables (6). The model developed

for k may not apply outside the range of vinasse

constituent concentration and temperatures (30-40°C)

tested in the experiments; additional work will need to be

conducted to verify whether this is the case. The

constituent values tested were generally low to moderate,

compared to values reported in the literature (Table 1),

with the exception of COD, which were very low to high.

In addition, the substrate used for model development was

glucose, which is likely easier for microorganisms to

digest than many of the compounds in real vinasse. This

led to a 20% over-prediction of the first-order rate

constant, in the limited validation that was conducted.

Ethics

Not applicable

REFERENCES

Albanez R, Chiaranda BC, Ferreira RG, Franca ALP, Honorio

CD, Rodriques JAD, Ratusznei SM, Zaiat M (2016).

Anaerobic Biological Treatment of Vinasse for

Environmental Compliance and Methane Production. Appl.

Biochem. Biotechnol., 178: 21-43. (Search Google Scholar ;

Import into EndNote )

Baez-Smith C (2006). Anaerobic Digestion of Vinasse for the

Production of Methane in the Sugar Cane Distillery, SPRI

Conference on Sugar Processing. (Search Google Scholar ;

Import into EndNote )

Barrera EL, Spanjers H, Romero O, Rosa E, Dewulf J (2014).

Characterization of the sulfate reduction process in the

anaerobic digestion of a very high strength and sulfate rich

vinasse. Chemical Engineering Journal, 248: 383-393.

(Search Google Scholar ; Import into EndNote )

Bose RC, Bush KA (1952). Orthogonal Arrays of Strength Two

and Three. Annals of Mathematical Statistics, 23: 508-524.

(Search Google Scholar ; Import into EndNote )

CONCLUSIONS AND RECOMMENDATIONS

Chen VCP (2001). Measuring the Goodness of Orthogonal

Array Discretizations for Stochastic Programming and

Stochastic Dynamic Programming. SIAM Journal on

Optimization, 12(2): 322–344. (Search Google Scholar ;

Import into EndNote )

Based on data collected from lab-scale reactors, a multiple

linear regression equation (R²= 0.80) was developed to

predict first-order methane generation rate constant k from

anaerobic digestion of vinasse. The selected best-fit model

for k varied positively as functions of temperature, initial

chemical oxygen demand, and the product of nitrogen and

phosphorous concentrations. The model predicted

methane generation from commercial vinasse within 20%.

The next step for future work is additional validation of

the model. Future work should include additional

experiments and model building at thermophilic

temperatures (50-60°C).

Cooper CD (2015). Introduction to Environmental Engineering,

Waveland Press, Long Grove, IL. (Search Google Scholar ;

Import into EndNote )

de Barros VG, Duda RM, Vantini JS, Omori WP, Ferro MIT, de

Oliveira RA (2017). Improved methane production from

sugarcane vinasse with filter cake in thermophilic UASB

reactors, with predominance of Methanothermobacter and

Methanosarcina archaea and Thermotogae bacteria.

Bioresource Technology, 244: 371–381. (Search Google

Scholar ; Import into EndNote )

To cite this paper: Kuusisto LM, Sattler ML, Chen VCP (2019). Predicting Bioenergy Potential from Vinasse Digestion: The VUMP Model (Vinasse Utilization for Methane

Production). J. Civil Eng. Urban., 9 (4): 36-42. www.ojceu.ir

Deublein D, Steinhauser A (2008). Biogas from Waste and

Renewable Resources, Wiley-VCH, Weinheim, 2008.

(Search Google Scholar ; Import into EndNote )

Kutner, MH, Neter J (2004). Applied Linear Statistical Models.

McGraw-Hill Education, 5th International edition. (Search

Google Scholar ; Import into EndNote )

dos Reis CM, Carosia MF, Sakamoto IK, Varesche MBA, Silva

EL (2015). Evaluation of hydrogen and methane production

from sugarcane vinasse in an anaerobic fluidized bed

reactor. International Journal of Hydrogen Energy, 40:

8498-8509. (Search Google Scholar ; Import into EndNote )

Kuusisto LM (2013). Development of a Mathematical Model,

VUMP (Vinasse Utilization for Methane Production). Ph.D.

Dissertation, University of Texas at Arlington, USA.

(Search Google Scholar ; Import into EndNote )

Moraes BS, Zaiat M, Bonomi A (2015). Anaerobic digestion of

vinasse from sugarcane ethanol production in Brazil:

Challenges and perspectives. Renewable and Sustainable

Energy Reviews, 44: 888-903. (Search Google Scholar ;

Import into EndNote )

España-Gamboa E, Mijangos-Cortes J, Barahona-Perez L,

Dominguez-Maldonado J, Herna´ndez-Zarate G, Liliana

Alzate-Gaviria L (2011). Vinasses: characterization and

treatments. Waste Management & Research, 29(12): 1235-

1250. (Search Google Scholar ; Import into EndNote )

Mota VT, Araujo TA, Amaral MCS (2015). Comparison of

Aerobic and Anaerobic Biodegradation of Sugarcane

Vinasse, Appl Biochem Biotechol., 176: 1402-1412.

(Search Google Scholar ; Import into EndNote )

España-Gamboa EI, Mijangos-Cortés JO, Hernández-Zárate G,

Domínguez Maldonado JA, Alzate-Gaviria LM (2012).

Methane production by treating vinasses from hydrous

ethanol using a modified UASB reactor. Biotechnology for

Biofuels, 5(82): 1-9. (Search Google Scholar ; Import into

EndNote )

Parnaudeau V, Condom N, Oliver R, Cazevieille P, Recous S

(2008). Vinasse organic matter quality and mineralization

potential, as influenced by raw material, fermentation and

concentration processes. Bioresource Technology, 99(6):

1553-1562. (Search Google Scholar ; Import into EndNote )

Espinoza-Escalante FM (2008). Estudio fisicoquímico

microbiano del proceso de digestión anaerobia de efluentes

agroindustriales orientado a la producción de hidrógeno y

metano. Tesis de Doctorado; Doctoral Dissertation,

Universidad de Guadalajara, Jalisco, Mexico.

Satyawali Y, Balakrishnan M (2008). Wastewater treatment in

molasses-based alcohol distilleries for COD and color

removal: A review. Journal of Environmental Management,

86(3): 481-497. (Search Google Scholar ; Import into

EndNote )

Fu SF, Xu XH, Dai M, Yuan XZ, Guoa RB (2017). Hydrogen

and methane production from vinasse using two-stage

anaerobic digestion. Process Safety and Environmental

Protection, 107: 81–86. (Search Google Scholar ; Import

into EndNote )

Shelton DR, Tiedje JM (1984). General Method for determining

anaerobic biodegradation potential. Appl. Envion.

Microbiol., 48: 840-848. (Search Google Scholar ; Import

into EndNote )

González LML, Reyes IP, Romero OR (2017). Anaerobic co-

digestion of sugarcane press mud with vinasse on methane

yield. Waste Management, 68: 139–145. (Search Google

Scholar ; Import into EndNote )

Syaichurrozi I, Budiyono, Sumardiono S (2013). Predicting

kinetic model of biogas production and biodegradability

organic materials: biogas production from vinasse at

variation of COD/N ratio. Bioresource Technology, 149:

390 – 397. (Search Google Scholar ; Import into EndNote )

Henze M, Harremoes

P (1983). Anaerobic treatment of

wastewater in fixed film reactors – a literature review.

Water Science and Technology, 15: 1-101. (Search Google

Scholar ; Import into EndNote )

US Department of Energy, Alternative Fuels Data Center (2016).

Global

Ethanol

Production,

Karanjekar R (2012). An improved model for predicting

methane emissions from landfills based on rainfall, ambient

temperature and waste composition. Ph.D. Dissertation, The

University of Texas at Arlington, Texas, USA. (Search

Google Scholar ; Import into EndNote )

http://www.afdc.energy.gov/data/10331 (accessed May

2016).

Vijayaraghavan K, Ramanujam RA (2000). Performance of

anaerobic contact filter in series for treating distillery

spentwash. Bioprocessing Engineering, 22: 109-114.

(Search Google Scholar ; Import into EndNote )

Khanal SK (2008). Anaerobic Biotechnology for Bioenergy

Production, Wiley-Blackwell. (Search Google Scholar ;

Import into EndNote )

Wilkie AC, Riedesal KJ, Owens JM (2000). Stillage

Characterization and anaerobic treatment of ethanol stillage

from conventional and cellulosic feedstocks. Biomass and

Bioenergy, 19: 63-12. (Search Google Scholar ; Import into

EndNote )

Kiyuna LSM, Fuess LT, Zaiat M (2017). Unraveling the

influence of the COD/sulfate ratio on organic matter

removal and methane production from the biodigestion of

sugarcane vinasse. Bioresource Technology, 232: 103–112.

(Search Google Scholar ; Import into EndNote )

Zieminski K, Kowalska-Wentel M (2015). Effect of enzymatic

pretreatment on anaerobic co-digestion of sugar beet pulp

silage and vinasse. Bioresource Technology, 180: 274-280.

(Search Google Scholar ; Import into EndNote )

Korndorfer GH, Anderson DL (1997). Use and impact of sugar

alscohol residues vinasse and filtercake on sugarcane

production. Sug y Azucar, 92: 26-35. (Search Google

Scholar ; Import into EndNote )

To cite this paper: Kuusisto LM, Sattler ML, Chen VCP (2019). Predicting Bioenergy Potential from Vinasse Digestion: The VUMP Model (Vinasse Utilization for Methane

Production). J. Civil Eng. Urban., 9 (4): 36-42. www.ojceu.ir