The anaerobic batch test (45 days at 37°C) was performed to describe the effect of thermal pretreatment at moderate temperatures (60, 80, and 100°C) over durations of 10 and 20 minutes on the enhancement of biogas production using hotel food waste from city of Jaipur, India. The results showed that the total cumulative biogas production with thermal pretreatment (100°C, 10 minutes) was 41% higher than the control. Also, this alternative gets first rank using multicriteria decision making model, VIKOR. This outcome was obtained due to the enhancement of degradation of organic compounds such as protein and volatile solids that occurred in the linear trend. Modified Gompertz and Logistic models were used to study the effect of different pretreatment parameters on lag time and biogas yield. Scanning electron microscopy and Fourier transform infrared spectroscopy were also employed to investigate the effect of thermal pretreatment on the physiochemical properties of food waste.
Together with the rapid growth of urban population and changes in the typical eating patterns, management of food waste (FW) has become an issue of international level [
Food waste percentage in different countries [
In India, the increase in foreign tourist arrivals has pushed the hospitality sector for enhancing quality services in order to fulfil customer’s satisfaction. This leads to the growth of hotels as well as FW generation, which has not been explored earlier. Rajasthan, which has a significant share of foreign tourism in the country with 7.2% Foreign Tourist Arrivals (FTA) in 2013 [
FW from hotels of Jaipur contains variety of materials, some of which are not suitable for anaerobic digestion (AD), namely, egg shells, coffee grounds, tissue papers, and bone, that have different physical and chemical characteristics. Due to presence of above undesirable components, it is not easy for microbes to degrade these complex and hard materials. Therefore, to improve the biodegradability of these materials and provide readily available organics for maximum biogas recovery, pretreatment becomes necessary prior to AD [
LHW pretreatment is one of the preferred pretreatment methods that can easily enhance the access of sugars to microbes by solubilizing them using higher temperature [
The degradation of fermentable sugars due to high severity of thermal pretreatment can be controlled by maintaining the optimum pH with the addition of base [
The impacts of thermal pretreatment on the chemical and physical properties of three municipal solid wastes (MSW), namely, kitchen waste, vegetable waste, and waste activated sludge, were explored by Liu et al. [
In spite of the advantages, thermal pretreatment has some disadvantages, i.e., (a) less stability, (b) formation of inhibitors due to side reactions, and (c) higher operating cost [
A series of batch experiments were conducted to measure the biogas potential and to compare the effect of thermal pretreatment using the hotel FW as the substrate. Both the increased hydrolysis and improved solubilization depends on temperature and time [
FW used in this study was collected on the weekly basis from the cafeteria and kitchens of the six hotels reputed with star category in Jaipur. After manual sorting, FW was separated from undesirable parts (such as egg shells, bones, tissue papers, peels, and kernels). Collected FW mainly consisted of carbohydrates, proteins, and fats, being the major components of rice, vegetables and fruits, bread, cooked pulses, and meat. All parts of collected food were mixed in a kitchen blender and stored at 4°C in a refrigerator after crushing into particles with an average size of 1-2 mm. Table
Analysis of the FW and inoculum.
| | |
---|---|---|
| ||
pH | 5.1 ± 0.2 | 7.8 |
Total solids (%) | 35.8 ± 0.5 | 6.4 ± 0.2 |
VS (%, wet basis) | 34.3 ± 0.4 | 4.4 ± 0.8 |
TCOD (mg/L) | 229.5 ± 38.7 | - |
SCOD (mg/L) | 79.3 ± 4.3 | - |
Protein (mg/L) | 20.3 ± 0.3 | - |
| ||
Carbon (%) | 49.5 ± 1.6 | 35.1 ± 1.2 |
Hydrogen (%) | 9.9 ± 0.2 | 4.3 ± 0.4 |
Nitrogen (%) | 2.7 ± 0.4 | 1.7 ± 0.2 |
Oxygen (%) | 36.1 ± 1.7 | 58.8 ± 0.9 |
Sulphur (%) | 0.3 ± 0.0042 | - |
C/N | 18.33 | 20.64 |
Inoculum was taken from the anaerobic digester plant (Rajasthan Gau Sewa Sangh, Durgapura, Jaipur, Rajasthan), which was using cow manure at the mesophilic temperature range, with initial pH of 7.8. Volatile solid (VS) concentration of the inoculum was 68.75% VS/TS. Before the experiment, it was stored in a container at room temperature for five days to acclimatize and starve it prior to using in AD batch experiments.
Thermal pretreatment was performed using hot water bath (Sanco, India) and 50g of a representative sample of FW (wet mass) and 50ml of distilled water to provide uniform heating for pretreatment making the solid to liquid ratio of 1:1 (w/v). The water bath was kept on until the target temperature reached to 60, 80, and 100°C in the beaker and held for the selected durations of time (10 and 20 min) for each temperature (Table
Severity factor and pretreatment conditions of hotel food waste.
| | |
---|---|---|
1 | 0.0 | Untreated |
2 | 1.7 | 60°C, 10 min |
3 | 2.0 | 60°C, 20 min |
4 | 2.3 | 80°C, 10 min |
5 | 2.6 | 80°C, 20 min |
6 | 2.9 | 100°C, 10 min |
7 | 3.2 | 100°C, 20 min |
To determine how the combined effect of temperature (T) and time (t) affects FW [
The FW digestion experiment was performed in the 610 mL serum bottle with active volume of 400mL (37±1°C, 90 rpm, 45 days) using orbital shaker (REMI CIS 24, India) to govern the biodegradability of untreated and thermally treated feedstock. Each bottle was fed with a mixture of FW and inoculum (with 1.2% TS after dilution) corresponding to the final concentration of 1.5g VS L−1 with initial pH between 7 and 7.3. Then, the upper space of each bottle was purged with nitrogen for at least 2 min and sealed by rubber plugs to guarantee anaerobic conditions. For each sample, the experiment was run in triplicate. Simultaneously, bottle containing inoculum alone was also made to measure the biogas generated from the inoculum. Initially the pressure (millibars) accumulated in bottle was measured by a pressure meter at the following time points: days 1, 2, 3, 4, 6, 9, 12, 15, 18, 22, 27, 32, and 45, which was further used to calculate the volume of biogas produced in each bottle using the following ideal gas law equation:
Biogas yield was calculated by taking the average of the biogas produced per VS of added substrate in triplicate bottles. After the pressure measurement, each bottle was depressurized by penetrating the needle into the rubber cap. The reported experimental results demonstrate the mean of triplicate made for each sample. Reported biogas yields from the substrates were calculated by subtracting the biogas production of the inoculum from the gross biogas production of the substrates.
To assess the performance parameter two models were used. The Modified Gompertz Equation (GM) (
Logistic function (LF) depends on the initial exponential increase that fits the global biogas production. This model assumes that the rate of biogas production is proportionate to the volume of gas already produced, the maximum production rate, and the maximum capacity of biogas production [
Lag phase (
Energy content in the generated biogas was calculated on the basis of methane content available in the gas (average, 62%), which was calculated using the ultimate analysis (Table
Gain/loss in the biogas energy (kJ/kg initial VS) was then measured by subtracting the biogas energy from the energy given for the hot water pretreatment.
The TS and VS content of FW was determined according to the standard methods [
Multicriteria decision making models (MCDM) are normally applied for both indefinite set of scenarios and definite set of scenarios. Pretreatment of FW is a definite set of scenarios having a definite set of output. For definite set of scenarios, there are many MCDM techniques such as ELECTRE (elimination et choix traduisant la realité), PROMETHEE (preference ranking organization method of enrichment evaluation), TOPSIS (technique for order preference by similarity to ideal solution), and VIKOR (VlseKriterijuska Optimizacija I Komoromisno Resenje) [
Create a decision matrix of alternative selected for experiment and output.
Create a normalized matrix using
After creating normalized matrix, find entropy of each alternative
Calculate dispersion value of each alternative
Find weight of each alternative
Determine utility measure (
Finally calculate VIKOR index,
All the data were tested for the level of significance and analysis of variance (ANOVA; p<0.05) was performed in Microsoft excel spreadsheet (version 2016) using solver function.
Figure
Variation in pH after thermal pretreatment of hotel food waste (two-factor ANOVA of data set showed p = 0.001662).
Thus, progressive increase in pretreatment severity resulted into decrease in pH, which may be due to the phase change of organic acids from solid to liquid in the FW. Other reasons could be the thermal hydrolysis reaction, i.e., high-pressure boiling of waste followed by rapid decompression, which leads to the reduction in pH. This process increases the biodegradability of waste along with reduction in the microbial load to an extent depending on temperatures and time.
The VS content for the untreated FW was 18.18 ± 1.33%. It can be concluded that, as pretreatment severity increases, no notable decrease in the VS content of the substrates was observed (Figure
Volatile solid proportion after thermal pretreatment in hotel food waste (two-factor ANOVA of data set showed p = 0.000025).
It is a parameter, which can be used as a performance indicator of digestion process in AD. It reflects the amount of soluble organic matter present in the substrates in the form of dissolved organic matter. Figure
Effect of pretreatment on SCOD content of hotel food waste (two-factor ANOVA of data set showed p = 0.03).
The obvious reason for the increased SCOD content on increasing the pretreatment temperature and duration is the breakage of chemical bonds, including VFA’s, polysaccharides, and proteins, by providing external energy in form of heat [
The results above support the liquid hot water pretreatment as the preferred choice for enhancing the digestion of FW. Kondusamy et al. [
Figure
Protein solubilization in hotel food waste (two-factor ANOVA of data set showed p = 0.0000009).
FT-IR analysis was performed in the range of 4000-400 cm−1 to characterise the effect of pretreatment on the functional groups and the chemical structure of untreated and pretreated FW (Figure
FTIR spectra of thermally pretreated hotel food waste.
SEM assists to understand the effect of thermal pretreatment on the microstructural properties of FW. Clusters of varying sizes and shapes were observed in the scanning electron micrographs, which were the composites of carbohydrates, proteins, and lipids (main organic compounds in FW). In contrast to treated FW, untreated FW contains more compact and cemented particles, which possess flat, rigid, and smooth surfaces of size ranging between 30 and 50
SEM images of FW under magnification of 1000x. (a) Untreated FW, (b) 60°, 10min, (c) 80°, 10min, and (d) 100°, 10min.
Therefore, it can be concluded that liquid hot water pretreatment resulted into decrystallisation of FW that would, in turn, be helpful in enhancing the contact between substrates and microorganisms via enhancing the available surface area for increased biogas production.
The improvement of cumulative biogas production from untreated and thermally pretreated FW is shown in Figure
Cumulative biogas production from untreated and thermally pretreated hotel food waste (two-factor ANOVA of data set showed p = 1.00647E-29).
The other curves at temperatures 60° and 80°C with treatment durations 10 and 20 min, respectively, demonstrated analogous trend due to linear increase in the SCOD and protein content, though the biogas production observed was less than the yield attained at 100°C, 10 min.
The development of refractory inhibitory compounds such as melanoidin and increase in concentration of soluble phenol at higher thermal pretreatment temperature and duration (100°C, 20 min) may reduce the activity of methanogens but not utterly after witnessing the high biogas yield compared to untreated FW [
The whole process of AD indicated the rapid degradation of feedstock and rather intensive biogas production [
Thus, the thermal pretreatment highlights the rapid solubilization of organic and inorganic macromolecules. Moreover, organic compounds with low molecular weight and increased surface area promote the contact between the microbes and substrate [
The expected kinetic parameters based on Gompertz and Logistic functions models are shown in Table
Results of kinetic study-modified Gompertz model and Logistic model.
| | | | | | | |
---|---|---|---|---|---|---|---|
Cumulative biogas production-experimental | 410 | 398 | 458 | 501 | 556 | 652 | 607 |
| |||||||
Cumulative biogas production-predicted | 410.31 | 396.98 | 452.10 | 482.45 | 541.98 | 627.64 | 553.22 |
Lag phase (days) | 0.5 | 0.06 | 0.05 | 0.03 | 0.02 | 0.01 | 0 |
R2 | 0.9909 | 0.9666 | 0.9528 | 0.9733 | 0.9837 | 0.978 | 0.9837 |
RMSE (%) | 5.95 | 2.95 | 2.07 | 2.34 | 2.56 | 2.02 | 1.97 |
| |||||||
Cumulative biogas production-predicted | 410.93 | 398.49 | 457.04 | 494.61 | 551.86 | 643.16 | 577.48 |
Lag phase (days) | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
R2 | 0.9937 | 0.9698 | 0.9573 | 0.9757 | 0.9875 | 0.9814 | 0.9852 |
RMSE (%) | 6.73 | 14.1 | 15.54 | 12.42 | 9.58 | 11.25 | 10.8 |
These two models (Gompertz and Logistic model) were also used by group of researchers ([
Biogas production potential of the untreated and treated FW at temperatures 60°, 80° and 100°C (each for 10, 20 min) were measured to be 555, 575, 645, 653, 692, 783 and 728 mL/g VS. Based on final CH4 content (average 62%) the methane content were 344, 357, 400, 405, 429, 485 and 451 mL/g VS respectively. The maximum biogas yield was obtained with the pretreatment temperature and duration of 100°C, 10 min. Lag time, RMSE and R2 values for both the models, at each operating temperature, are shown in Table
Calculated lag time for both the kinetic models was found to be nearly zero for every case because of the presence of active bacteria in the added inoculum and readily accessible biodegradable component in the FW. It also signifies the rapid consumption of soluble material by anaerobic biomass [
Untreated FW can generate 12.89 MJ/kg (VS basis) of energy via AD. It was expected that pretreatment can improve the net benefit by increasing the biogas yield. However, pretreatment can cause dry matter loss and requires additional energy, which can offset the improvement in biogas yield. As shown in Table
Net benefit of energy production from biogas due to the thermal pretreatment of FW.
| | | | |
---|---|---|---|---|
60°, 10min | 1.7 | 13.35 | 46.06 | Negative |
80°, 10min | 2.3 | 15.15 | 76.66 | Negative |
100°, 10min | 2.9 | 18.16 | 107.46 | Negative |
Untreated FW | 0 | 12.89 | 0 | 12.89 |
After the analysis was perfomed and lots of data were genrated; for multicriteria analysis of thermal pretreatment of FW, a set of output data was selected for employing VIKOR method. A set of possible alternatives based on number of tretament temperature and time duration was prepared for getting best tretament condition (Table
Decision matrix as per experimental results of thermal pretreatment of FW.
| | | | | |
---|---|---|---|---|---|
60°C, 10 minutes (A1) | 5.69 | 17.09 | 83.85 | 21.6 | 575.66 |
60°C, 20 minutes (A2) | 5.29 | 16.87 | 88.62 | 21.8 | 645.73 |
80°C, 10 minutes (A3) | 5.66 | 16.66 | 91.36 | 22.6 | 653.24 |
80°C, 20 minutes (A4) | 5.19 | 16.49 | 92.45 | 22.9 | 692.68 |
100°C, 10 minutes (A5) | 5.58 | 16.26 | 101.56 | 23.3 | 783.23 |
100°C, 20 minutes (A6) | 4.96 | 16.09 | 100.12 | 23.7 | 728.48 |
Sum | 32.37 | 99.46 | 557.96 | 135.9 | 4079.02 |
Max | 5.69 | 17.09 | 101.56 | 23.7 | 783.23 |
Min | 4.96 | 16.09 | 83.85 | 21.6 | 575.66 |
Max–min | 0.73 | 1 | 17.71 | 2.1 | 207.57 |
In Table
Normalized matrix is obtained as below for each alternative, criterion using (
| | | | | |
---|---|---|---|---|---|
A1 | 0.175780043 | 0.171827871 | 0.15027959 | 0.158940397 | 0.141127035 |
A2 | 0.163422922 | 0.169615926 | 0.15882859 | 0.160412068 | 0.158305181 |
A3 | 0.174853259 | 0.167504524 | 0.163739336 | 0.166298749 | 0.16014631 |
A4 | 0.160333642 | 0.165795295 | 0.165692881 | 0.168506255 | 0.169815299 |
A5 | 0.172381835 | 0.163482807 | 0.182020217 | 0.171449595 | 0.192014258 |
A6 | 0.153228298 | 0.161773577 | 0.179439386 | 0.174392936 | 0.178591917 |
After making each entry of the data set dimensionless, entropy value, dispersion value and weight of each alternative was determined (Table
Determining the value of entropy, dispersion, and weight of each alternative using (
| | | |
---|---|---|---|
A1 | 0.999302322 | 0.1842 | 0.18328732 |
A2 | 0.999302322 | 0.177041 | 0.176163955 |
A3 | 0.999302322 | 0.167215 | 0.166386118 |
A4 | 0.999302322 | 0.168169 | 0.167336036 |
A5 | 0.999302322 | 0.146966 | 0.146238199 |
A6 | 0.999302322 | 0.161388 | 0.160588372 |
After getting the values of entropy, dispersion and weight, utility measures, regret measure and VIKOR index were calculated (Table
Utility measure (
| | | | | | | | |
---|---|---|---|---|---|---|---|---|
A1 | 0 | 0 | 0.248051576 | 0.062278006 | 0.525738008 | 0.83606759 | 0.525738008 | 1 |
A2 | 0.076771049 | 0.005241554 | 0.181241524 | 0.056346768 | 0.348263122 | 0.667864016 | 0.348263122 | 0.716785979 |
A3 | 0.005757829 | 0.010244856 | 0.142864262 | 0.032621813 | 0.329241623 | 0.520730382 | 0.329241623 | 0.604021944 |
A4 | 0.095963811 | 0.014295148 | 0.127597394 | 0.023724955 | 0.229347096 | 0.490928403 | 0.229347096 | 0.486020253 |
A5 | 0.021112038 | 0.019774954 | 0 | 0.011862477 | 0 | 0.05274947 | 0.021112038 | 0 |
A6 | 0.140107164 | 0.023825246 | 0.020169072 | 0 | 0.138672043 | 0.322773525 | 0.140107164 | 0.290263409 |
Finally, rank of alternatives were provided as per VIKOR index obtained in ascending order, and best alternative is one having minimum value of VIKOR index (Table
Rank of alternatives.
| | |
---|---|---|
60°C, 10 minutes (A1) | 1 | 6 |
60°C, 20 minutes (A2) | 0.716785979 | 5 |
80°C, 10 minutes (A3) | 0.604021944 | 4 |
80°C, 20 minutes (A4) | 0.486020253 | 3 |
100°C, 10 minutes (A5) | 0 | 1 |
100°C, 20 minutes (A6) | 0.290263409 | 2 |
The obtained rank showed the effect of pretreatment and time duration on various factors tabulated in Table
According to MCDM using VIKOR technique, pretreatment of FW at 100°C, 10 min was observed to be the best treatment condition among others and also experimentally this alternative helps to achieve maximum biogas yield (783 ml/g VS).
Thermal energy was employed for FW pretreatment concerning augmentation of biogas yield in the AD process. A significant enhancement of organic matter solubilization and biogas production from FW was observed after thermal pretreatment. A direct correlation between soluble COD, soluble protein and biogas production was observed. However, the soluble COD did not increase after the temperature and duration of (100°C and 10 min). Thermal pretreatment enhances the degradation of organic and inorganic compounds, which leads to efficient AD of FW treated at high temperature and longer duration. SEM analysis showed that the thermal pretreatment disrupts and reduces the size of the food particles and increases the roughness to promote the contact between the substrate and microbes. Whereas, FTIR also showed the presence of carbohydrates, proteins and lipids as well as their conversion in to simpler forms with the increasing severity of pretreatment. Both kinetics models (Modified Gompertz and Logistic function) showed agreement with the experimental curve and fit up to the similar extent with R2 value greater than 95%. Net energy analysis showed that thermal pretreatment was not an economical method of pretreatment as it incurs more energy as input than additional output energy from biogas. Therefore, in the present form it may have less feasibility; however employment of residual CHP heat for FW pretreatment may lead to development of an energy efficient, sustainable, and economic process.
The authors alone are responsible for the content and writing of the paper.
The authors declare that they have no conflicts of interest.
Paras Gandhi and Kunwar Paritosh acknowledge fellowships from Malaviya National Institute of Technology Jaipur and facilities at Centre for Energy and Environment, Malaviya National Institute of Technology, Jaipur India. Javier Lizasoain and Andreas Gronauer thank Department of Sustainable Agricultural Systems, University of Natural Resources and Life Sciences, Vienna, Austria, for facilities. Vivekanand Vivekanand and Alexander Bauer are thankful to Department of Science and Technology (Grant no. INT/AUSTRIA/BMWF/P-01/2017) Government of India and WTZ-BMWF Austria for financial support, respectively.