INTRODUCTION
What are Fermentation Tests Required for?
The economic efficiency of biogas plants is largely determined by the substrate costs. Approx. 40% to 60% of the annual costs of a biogas plant, fed with energy crops, are attributable to substrate procurement. It is, therefore, essential to calculate standard figures when planning a biogas plant. When using liquid manure and by-products, it is also essential to know the specific methane yield of the substrates used in order to enable optimum planning of a biogas plant. In addition to the data on the specific methane yield, further information on the fermentation substrate used is also required. In particular, the yield per unit area of fresh matter, the dry matter yield and the quality of the fodder have a considerable influence on the amount of methane that can be achieved.
In principle, the content of carbohydrates, fats and proteins correlates directly with the energy content of the substrate used and should actually be sufficient as a basis for calculating the methane yield. There are also a number of methods for this estimation. Starting with the calculation by Buswell (1939), who estimated the biogas yield, and the methane and carbon dioxide content in the biogas on the basis of the elementary composition of the substrate. Other calculations were made to determine the nutrient composition by Baserga (1998) and later by Keymer & Schilcher (1999). These values can only be used as a rough guide for simple planning of full scale biogas plants. As Czepuck et al. (2006), proved in comparison studies, the results deviate by 10 % to 20 % from the measured value in the biogas yield test. Since the plant substrate mass is composed of very different nutrient combinations, further estimation formulas have been proposed by Amon et al. (2007) and Kaiser (2007), as well as the laboratory determination of the fermentable volatile solid (FVS) value. VDI 4630 (2016) refers that some of these indicators allow a limited extension of the reliability of the results.
FIGURE 1.
Methane yields and formation kinetics of different nutrient fractions (according to Czepuck et al., 2006)
Above all, the different compositions of the carbohydrate fraction and its extremely different digestibility have a clear effect on the degradability, the degradation kinetics and finally the achievable gas yield. While starch can be completely converted into biogas in about 5 days depending on its storage in the plant, the conversion of plant supporting tissues such as hemicellulose and cellulose is significantly slower. Especially if a high proportion of lignocellulose complexes is present, this has a negative effect on the methane yield and the substrate degradability. Therefore, a fermentation test is the most exact and recommended method for the exact determination of methane yield of plant mass (Czepuck et al., 2006).
DEVELOPMENT
Reference on the Methane Yield of Common Substrates
For the use of common substrates, German KTBL (Advisory Board for Technology and Construction in Agriculture) has considered a great amount of research, which has been compiled with the help of a large number of individual results from several well-known laboratories and research institutions (KTBL, 2011). As a rule, these reference values are extremely valuable and very useful for the design and planning of biogas plants.
However, it must be noted that depending on the location, variety, vegetation course during the year, harvest time, storage and silage, significant deviations from the standard values are possible. For this reason, laboratory analysis in the biogas laboratory is useful in many cases for determining methane yield.
Guidelines for Fermentation Trials
Many guidelines are available to get a higher reproducibility of Biomass Methane Potential (BMP). Beside European standards (Holliger et al., 2016), VDI 4630 is the most common guideline used in Germany. Guideline 4630 (VDI 4630, 2016) describes both, the procedure for carrying out batch and continue tests. KTBL "Interlaboratory Tests" working group developed, together with the VDLUFA, the VDLUFA Method for "Determination of biogas and methane yields in fermentation tests" VDI 4630 (2016), which somewhat simplifies the most important criteria for carrying out fermentation tests. The aim of this method is to improve the quality assurance of laboratory tests on biogas yield.
The regulations stipulate that a minimum number of test conditions must be satisfied:
Use of suitable, gas-tight and tempered "small fermenters"
Use of a suitable inoculum (either specially bred or fermentable material from active biogas fermenters; inoculum);
Mixing ratio of inoculum and fermentation substrate (volatile solid (VS)-related) in a ratio of at least 2:1 to find sufficient buffer capacity in the fermentation Batch;
Inoculum ferment in parallel as a zero sample;
At least one standard substrate is also fermented (e.g. microcrystalline cellulose) in internal laboratory conditions;
Fermentation temperature should be 37°C ± 1°C;
Measure biogas formation as often as possible and determine the methane content with each gas extraction;
At least 25 days of fermentation time. Termination criterion for the test batch: if less than 0.5 % of the gas quantity formed to date is produced on at least 3 consecutive days;
Reference of the methane yield to the VS input in the test substrate;
If silages or substrates with volatile components (fermentation acids, alcohols) are used, the value of the VS content must be corrected (Mukengele & Oechsner, 2007);
Standardization of gas production (0°C, 1013 hPa); consideration of water vapor.
In the evaluation, the inoculum own production of biogas/methane is deducted from the total production in order to determine only the biogas/methane yield of the sample tested. The amount of biogas/methane produced over the experimental time is shown as a sum curve.
Comparisons among Laboratories to Ensure Test Quality
Now in Germany, at least 40 laboratories offer tests to determine the methane yield of substrate samples. These laboratories are very experienced and they are involved in reducing the risk of errors due to the complex and multi-stage test procedures. In previous publications, the methane yields differ sometimes significantly from each other.
For this reason, KTBL working group has established a system of comparisons among laboratories of biogas in cooperation with VDLUFA-NIRS GmbH and with the financial support of BMELV. So far, eleven tests have taken place. The number of participating laboratories has been in the range of 20 to 33 laboratories. By evaluating the test results and a comprehensive description of the method as well as a detailed error analysis, the participants were able to identify and eliminate errors in their own procedures. All participating laboratories were required to comply with the VDI Guideline 4630 and the VDLUFA method regulation. Microcrystalline cellulose was used as the standard fermentation substrate for each pass. Besides that, other fermentation substrates were chose. These substrates should cover the usual range of substrate variations from practice. Identical sample material was sent to all laboratories in all rounds. The samples were crushed. The substrates valued were wheat grain, dried maize, dried grass, maize silage and grass silage and rape press cake. When fresh silages were shipped, the effects of sample storage and sample homogenization on the final result were also possible to investigate.
It was noticeable, in the first run in 2006, the results for cellulose showed a relatively wide dispersion, although it was a very homogeneous and standardized substrate. The comparison of methane yield among the different laboratories show a variation coefficient between 19.5% and 8.4% in the eleven tests analyzed. When comparing the test setup and the results, it became clear that the differences were not due to the type and size of the respective test facilities, but to the accuracy of methane measuring instruments, their regular calibration, the mathematical evaluation under consideration of the reference variables for standard conditions and the consideration of water vapor correction in the event of deviations, which played a more relevant role.
Influence of the Mode of Operation - Transfer from Batch to Continuous Operation
With the exception of a few systems, biogas plants are not operated as batch type in practice. Instead of this, they operate as continuous systems with addition of fresh fermentation substrate several times a day. As a result, the process steps for anaerobic decomposition of organic matter run parallel and simultaneously. While the batch approach shows a clear change in biogas composition during the test sequence, the continuous approach hardly allows a difference in gas quality to be measured over the course of the day, if the same substrate is fed continuously. In addition, with fully mixed fermenters, but also with the so-called "plug flow fermenters", in close correlation with the hydraulic retention time of the substrate in the fermenter as well as with the fermenter geometry and the feeding frequency, certain parts of fresh or only partially degraded substrate, always leave the fermenter before the hydraulic retention time has elapsed. This occurs even before their entire methane yield potential has been obtained in the substrates.
In batch systems, on the other hand, it is waited until the fermentation process has almost completely decayed. For this reason, deviations are to be expected when batch results are transferred to continuous operation. Today, biogas plants in Germany are equipped with very long hydraulic retention times, partly due to legal requirements and also due to many of them operate in cascade arrangement to achieve the methane yield potential be fully exploited. Methane potentials and methane yields per substrate could then be determined at the same time in the experiments. By determining the residual gas potential as in the overflow of biogas fermenters, it is relatively easy to check to what extent the fermentation substrate used is utilized or whether there are still reserves in its potential energy. Investigations in various biogas plants in production as well as in plants in cascades have shown a narrow correlation between the gas potential of substrates and the hydraulic retention time (Mönch, 2014). As a rule, only a very low residual gas potential (< 5 % when determining the residual gas potential at mesophilic temperature) was measured in the retention time of the fermenter (including the secondary fermenter) over 100 days. The results of two federal measurement programs (FNR, 2005; 2009), confirm these statements. Research carried out at 25 biogas plants with different substrates and different cascade digesters showed, that the transferability from laboratory results to real plants is possible (Ruile et al., 2015). In case of comparison of batch to full-scale biogas plant, it seems like the influence of low data quality on the full-scale biogas plant results in a bigger mistake than the measurement error based on batch tests.
In an extensive trial at the University of Hohenheim, various fermentation substrates were investigated in a combination of batch tests and continuous trials in the biogas laboratory. For continuous trails the residual gas potential of the fermenter in overflow was measured in batch tests. In these tests, maize silage, ground wheat seed and mixtures of both substrates were fermented. The substrates were dried at 60 °C and ground to inhibit possible effects of these pretreatments, then they were fermented in the digesters. The substrates were supplemented with liquid manure (17 % VS content) to enable stable operating conditions. Liquid manure was also fermented alone as a control variant. In continuous operation, 15 horizontal fermenters each with a useful fermenter volume (FV) of 17 l, were used with two organic loading rates (2.5 and 4.0 kg VS m-³ FV d-1). The hydraulic retention time in the continuous fermenters was 35 days at a fermentation temperature of 37 ± 2°C. For each variation two repetitions were applied and the experiments lasted over 123 days, (= more than 3 retention times). In addition to the amount of biogas and methane, the volatile fatty acid content and the FOS/TAC value were regularly monitored.
The fermenter operation was very stable in all variants investigated. Only the maize silage variant with an organic loading rate of 4 kg VS/m³FVd showed an increase in the fatty acids (HAC) to a maximum of 6,500 mg/l and a slight drop in the pH value to 7.2 from the 67th day of the experiment onwards. All other variants were stable (pH values 7.4 to 7.6) (Mukengele, 2017).
The discharge (fermentation residue) of the continuous running fermenters was collected towards the end of the experiment (105th, 108th and 115th days) in order to determine the residual methane potential in the Hohenheim biogas yield test (HBT) at a fermentation temperature of 37 °C for 35 days. In addition, the methane yield of fodders used as fermentation substrates was determined as standard in the HBT at 37 °C and a retention time of 35 days. This resulted in relatively high methane yields between 0.377 and 0.399 Nm³ methane per kg VS. Values between 4.7 and 5.3 kWh/kg VS were measured with the bomb calorimetric bomb.
In Figure 2 a balance of the various experimental approaches is shown. It can be seen, that with a high volume load (OLR 4) in continuous running digesters, especially with maize silage, less methane yield (81.0 % of the HBT potential) can be achieved than with a low loading rate (89.5 % with OLR 2.5). This also tends to be the case for cereals, but due to the very good degradability of ground cereal grain this is of little significance (89.7 % and 91.2 %).
FIGURE 2.
Methane yields from a continuous test at different OLR with subsequent determination of the residual methane potential in comparison to the methane yields from batch tests using the HBT method - comparative presentation.
The potential of residual gas and methane yields of the continuous tests resulted near to the values obtained in batch tests (values between 98.7 and 101.1 %). This proves that the results from batch tests can be transferred with high precision to continuous tests. However, it is essential to ensure that part of the methane potential that is flushed out of the fermenter via the fermentation residue, particularly in short retention times, (35 days), is quantified and utilized. It is not like that in cases of substrate hydraulic retention times of more than 100 days and several fermenters are connected in cascade.
Transferability of the Results in Batch Test to Plants in Production
Possible causes for deviations in the gas yields between laboratory and practice are the different fermentation conditions, e.g. the rate of organic loading or the "unpredictable" influences of biological processes in real practice. An overview of the differences between batch tests and real biogas plants is shown in Table 1 below.
In the design of biogas plants, the ratio of fermenter size and volume of organic load planned of the CHP plays an important role. In biogas plant management, the daily supply of substrates through the substrates used and their gas yields is decisive. Some uncertainties still exist related to fermentation tests, their evaluation and transferability to practical plants, mainly referred to biogas production (KTBL, 2011).
Possible deviations from the guideline values can result from the substrates, as they are required in large quantities for the biogas plant and their substrate properties can vary depending on the variety, harvest time and year of cultivation. For example, the substrate also changes its composition (DM/VS content, content of fermentation acids, pH value) in the bunker silos over the storage period. The determination of the substrate input quantity is difficult despite existing weighing equipment, because the scales often do not have the required accuracy, between the mixing vessel and the input screws and because the individual substrates are often not recorded separately and exactly, specially, if mixed silage is used as substrate. Neither the amount of liquid manure nor other liquids (rainwater, silo leachate) are measured.
TABLE 1.
Selected process differences between batch test setups and biogas plants
In the fermenter, process-related factors such as the mixture of substrates, their content of nutrients (especially trace nutrients), the biological environment in the fermenter, the retention time and the load volume also have an effect on the methane yield.
As with the implementation of measurement programs, like Federal Measurement Programs I and II. (FNR, 2005; 2009), there are in practice considerable problems in determining the biogas yield and especially the biogas quality. The electricity yield is often measured and estimated via the electrical efficiency of the CHP due to a lack of gas meters. The actual proportion of ignition oil and possibly own electricity consumers running through the meters must also be taken into account. In most cases, no conversion of standard conditions is carried out when providing practical measurement data. This can cause an overestimation of the gas volume over 20 % (Ruile et al., 2015).
Practical Data and Batch Tests
The University of Hohenheim has a research biogas plant on a practical scale with an output of 350 kW-h/day. This plant is equipped with two separate fermenter lines (800 m³ usable volume each), so that comparative investigations are possible. The plant is also high equipped. The measurement devices are also frequently calibrated. It guarantees a high quality of data. In a study carried out by Mönch (2014), a comparison was made between the data at this plant and those on a practical scale, similar to what had previously been made with data obtained at laboratory scale. This biogas plant is intensively monitored, the quantity and quality of all input materials and the results obtained are precisely recorded and, as far as technically possible, kept constant during all the process. In the experiment, a relatively high proportion of liquid manure (50.0 %) was used (fresh mass). In addition, horse manure with other solid manure (23.6 %), maize silage (10.1 %), grass silage (8.8 %), cereal whole plant silage (4.6 %) and ground cereal grain (2.9 %) were used. The VS share of horse manure was 27.5 %. The substrate quality of all input materials was determined and as well as biogas and methane yields obtained in fermentation tests. The fermenters are equipped with a mechanical processing technology (Cross flow grinder, MEBA, Nördlingen, Germany). The daily input quantity of fresh mass was 12.1 t/d with a standard deviation of 2.9 %. The organic loading rate was 2.49 kg VS per m³ of FV/day and the HRT was 62.4 days.
With the aid of the weighed input quantities, their content of volatile solids and the laboratory values for the methane yield, the expected methane quantity of the fermenter was calculated and compared with the values measured by the gas meter of the fermenter. The results are shown in Figure 3.
FIGURE 3.
Comparison of the methane production determined by input quantity and HBT batch measurement with the actually measured methane production of the fermenter (Mönch, 2014).
During the 20-days observation phase there were certain fluctuations in the measured values, which were more variable in the input quantities. This dispersion can be explained by the fact that the daily fluctuations of the VS input have a direct effect on the calculated value for methane formation and are thus clearly visible in the graph, while these fluctuations in the measured values are compensated by the substrate degradation in the fermenters while several days passed. The mean specific methane yield of the values calculated via HBT was on average 305 l CH4/kgVS respect to the values measured in the fermenters, 311 l CH4/kgVS, only 2 % higher. This very small deviation between the methane yields determined from laboratory values and the measured values at the practical fermenter confirms that the laboratory values can be used very well for an estimate of the methane yield to be expected and thus for an economic efficiency estimate. This applies if a representative sample was analyzed for the methane yield; the VS content of the substrate input and its exact weight were recorded regularly and at short intervals. As a rule, as this last example has shown, inaccuracies are more likely to occur when determining the mass in practice than when determining the methane yield in the laboratory.
RESULTS AND DISCUSION
The work shows the feasibility of extrapolating results obtained to laboratory scale to real scale in the aspects related to the design and exploitation of biogas plants. These techniques of simulation of processes to small scale are much utilized at international level. In the Cuban case, this constitutes an obligatory reference and their applicability saves time, resources and money. In the last decade, works based on the simulation of fermentation processes to small scale in this thematic have been developed by investigators at Central University "Marta Abreu" of Las Villas and by others at University "José Martí Pérez" of Sancti Spíritus, where investigation projects, master and Doctorate thesis have being developed in theme analyzed in this work.
CONCLUSIONS
The determination of methane yield by batch tests is now widely used in Germany and is carried out by at least 40 laboratories. Their quality is regularly assured by inter laboratory comparisons.
Examples in the laboratory and in practice have shown that there is a relatively good correlation between laboratory and practical values for the methane yield. This also proved that the gas yield guidelines compiled by the KTBL on the basis of laboratory values from several well-known biogas laboratories are important and indispensable for the economic preliminary planning of biogas plants.
In some cases, there are certain deviations from the assumed values at the biogas plant operated later. However, a large number of influencing factors affect the fermentation substrate used and its quality, but are also linked to the process and operating mode of the biogas plant.
The production of biogas is a microbial degradation process involving a large number of microorganisms. Here, certain deviations in the range of 5 to 10 % are always possible. In order to ensure the accuracy of the laboratory tests when determining the methane yield, regular participation in inter laboratory comparisons and constant internal laboratory testing using standard substrates should be a way of validating the results obtained.