Journal of Civil Engineering and Urbanism  
Volume 9, Issue 4: 36-42; July 25, 2019  
ISSN-2252-0430  
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  
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  
100 billion L of ethanol were produced in 2015. This  
generated 900 to 2000 billion L of vinasse (US  
Traditionally, Brazil, Australia, and other countries  
have disposed of vinasse by applying it as a fertilizer on  
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.,  
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 6798% (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  
CcHhOoNnSs + yH2O xCH4 + (c-x)CO2 + nNH3 + sH2S (1)  
where  
x = 1/8 * (4c+h -20 3n2s) and y = ¼ * (4ch20+3n+3s).  
A variety of previous studies have measured  
methane generation from anaerobic treatment of vinasse  
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  
36  
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  
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  
a
wider range of vinasse  
compositions. An experimental design (strength  
2
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  
37  
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)  
K
S (as SO4)  
1 - Low COD  
2,600  
75,000  
60  
1200  
7
39  
39  
34  
580  
2 Medium COD, Low P, K  
3 High COD  
7
147,000  
550  
90  
90  
39  
580  
4 Medium COD, High P, K  
Range of values tested  
Literature values a,b  
75,000  
1200  
1742  
34  
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  
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),  
Lo = 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 22444XR), 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  
t = time (days).  
Rearranging Eq. 1 and taking the natural log of both  
sides gives:  
ln(1-V/Lo) = -kt  
(2)  
If ln(1-V/Lo) 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. Lo was estimated from the horizontal  
asymptote of the plots of ln(1-V/Lo) vs. time. When the  
plot did not clearly reach an asymptote, the value of Lo  
was chosen which gave the largest R2 value for a  
regression line fit to ln(1-V/Lo) 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 = Lo(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  
38  
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 R2 value, adjusted R2 value, Mallows Cp, and  
Akaike Information Criterion (AIC) or Schwarz Bayesian  
Criterion (SBC). Parsimonious models with high R2 and  
adjusted R2, low Mallows Cp, 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  
39  
The R2 values for the other curve fits were above 0.75 for  
all reactors.  
30°C  
35°C  
40°C  
50  
40  
30  
20  
10  
0
Table 2. Rate constant k values for the anaerobic reactors  
Temp.  
(0C)  
30  
k
Composition  
(/day)  
N/A  
1
1
1
2
2
2
3
3
3
4
4
4
35  
40  
30  
35  
40  
30  
35  
40  
30  
35  
40  
N/A  
0.67  
0.52  
0.61  
0.62  
0.70  
0.74  
0.88  
0.56  
0.57  
0.79  
0
2
4
Day  
6
8
a) Composition with initial low COD  
30°C  
35°C  
40°C  
50  
40  
30  
20  
10  
0
0
2
4
6
8
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  
60  
50  
40  
30  
20  
10  
0
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)  
0
2
4
6
8
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  
kcalculated value 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, kactual was 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  
50  
40  
30  
20  
10  
0
0
2
4
Day  
6
8
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  
40  
methane in anaerobic digestion. The temperature  
dependence is not surprising either, given that microbial  
activity typically varies as a function of temperature  
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.  
Bose RC, Bush KA (1952). Orthogonal Arrays of Strength Two  
and Three. Annals of Mathematical Statistics, 23: 508-524.  
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): 322344. (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: 371381. (Search Google  
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  
41  
Deublein D, Steinhauser A (2008). Biogas from Waste and  
Renewable Resources, Wiley-VCH, Weinheim, 2008.  
Kutner, MH, Neter J (2004). Applied Linear Statistical Models.  
McGraw-Hill Education, 5th International edition. (Search  
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:  
Kuusisto LM (2013). Development of a Mathematical Model,  
VUMP (Vinasse Utilization for Methane Production). Ph.D.  
Dissertation, University of Texas at Arlington, USA.  
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-  
Mota VT, Araujo TA, Amaral MCS (2015). Comparison of  
Aerobic and Anaerobic Biodegradation of Sugarcane  
Vinasse, Appl Biochem Biotechol., 176: 1402-1412.  
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  
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):  
Espinoza-Escalante FM (2008). Estudio fisicoquímico  
y
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  
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: 8186. (Search Google Scholar; Import  
Shelton DR, Tiedje JM (1984). General Method for determining  
anaerobic biodegradation potential. Appl. Envion.  
Microbiol., 48: 840-848. (Search Google Scholar; Import  
González LML, Reyes IP, Romero OR (2017). Anaerobic co-  
digestion of sugarcane press mud with vinasse on methane  
yield. Waste Management, 68: 139145. (Search Google  
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  
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  
2016).  
Vijayaraghavan K, Ramanujam RA (2000). Performance of  
anaerobic contact filter in series for treating distillery  
spentwash. Bioprocessing Engineering, 22: 109-114.  
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  
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: 103112.  
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.  
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  
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  
42