- Email: [email protected]
Bioavailability of essential trace elements and their impact on anaerobic digestion of slaughterhouse waste
Biochemical Engineering Journal 99 (2015) 107–113
Contents lists available at ScienceDirect
Biochemical Engineering Journal journal homepage: www.elsevier.com/locate/bej
Bioavailability of essential trace elements and their impact on anaerobic digestion of slaughterhouse waste Markus Ortner a,b,∗ , Michael Rameder a,1 , Lydia Rachbauer b,1 , Günther Bochmann b,2 , Werner Fuchs a,3 a b
University of Natural Resources and Life Sciences, IFA Tulln, Institute for Environmental Biotechnology, Konrad Lorenz-Str. 20, A-3430 Tulln, Austria Bioenergy 2020+, Konrad Lorenz Str. 20, A-3430 Tulln, Austria
a r t i c l e
i n f o
Article history: Received 14 January 2015 Received in revised form 2 March 2015 Accepted 24 March 2015 Available online 25 March 2015 Keywords: Anaerobic processes Anaerobic digestion Trace elements Sequential extraction Chemical speciation Bioavailability Biogas Waste treatment Bioprocess monitoring Slaughterhouse waste
a b s t r a c t Slaughterhouse waste is an energy rich feedstock suitable for anaerobic digestion processes. However, chemical characterization showed a deficiency in essential trace elements which are critical for optimal performance of the process. Hence this study investigated the degree of bioavailability of trace elements in four semi-continuous lab-scale AD tests accepting slaughterhouse waste under mesophilic conditions (38 ◦ C) and a moderate organic loading rate of 2.2 kg/m3 d. Parameters, such as volatile fatty acid (VFA) concentration, COD removal rate and specific methane yield were compared to the results of sequential extraction analysis. The highest methane yield (250–275 Nm3 /t COD), lowest accumulation of VFA (<500 mg/l) and high COD removal rate (75–80%) was obtained when the total concentration of 11.4 mg/l Ni, 25.4 mg/l Co and 4.8 mg/l Mo was present in the reactor, of which 62% of Ni and Co, and 68% of Mo were bioavailable for microbial uptake. Based on these results it can be recommended that a supply of 2.5 g/t Ni, 3.5 g/t Co, 0.6 g/t Mo and 0.05 g/t Se provide optimal conditions for anaerobic digestion of slaughterhouse waste. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Optimal supply of trace elements is a prerequisite for microbial growth and metabolism in anaerobic digestion (AD) processes. Consequently, deficiency causes limitation of the activity of the microbial consortium [1]. This is because, methanogenesis, the final metabolic pathway during AD, involves the participation of various metal-rich enzymes such as carbon monoxide dehydrogenase/acetyl-CoA synthase (Cdh) or methyl coenzyme M reductase (Mcr). These enzymes catalyse key metabolic steps and require sufficient supply of Fe, Ni and Co [2,3]. The exact amount of required trace metals of AD may vary depending on the involved microbial species and their methanogenic pathway, but there are
∗ Corresponding author. Tel.: +43 2272 66280 536. E-mail addresses: [email protected] (M. Ortner), [email protected] (M. Rameder), [email protected] (L. Rachbauer), [email protected] (G. Bochmann), [email protected] (W. Fuchs). 1 Tel.: +43 2272 66280 535. 2 Tel.: +43 2272 66280 536. 3 Tel.: +43 2272 66280 553. http://dx.doi.org/10.1016/j.bej.2015.03.021 1369-703X/© 2015 Elsevier B.V. All rights reserved.
some general tendencies. Fe is the most abundant element, followed by Ni and Co, and smaller amounts of Mo, W and Zn. Fe is part of Fe–S clusters, which are used for both, electron transport and/or catalysis. Ni is either bound to Fe–S clusters or in the active centre of the co-factor F430. Co is present in cobamides involved in methyl group transfer. Mo and W are non-covalent bound to co-factors molybdopterin and tungstopterin, which catalyse two electron redox reactions (reduction of CO2 to formate by formate dehydrogenase (FDH)) [3]. Furthermore, Mo also plays a crucial role in the syntrophic propionate oxidation [4]. In numerous research studies the positive impact of trace element supplementation on anaerobic digestion processes was demonstrated. It was reported that productivity, bio-methanation rate and hence process stability were significantly improved. However, uncertainty about the exact concentration of essential elements that need to be present in a well operating bio-process remains. Suggested values cover a wide spectrum of concentrations comprising a range of more than three decimal powers [5–7]. A major factor of influence is that presence of high metal concentrations does not necessarily imply that microorganisms are also able to take them up and incorporate them into catalytic centres of their enzymes [8]. Microorganisms are using two ways for metal uptake, a fast, passive and unspecific one, driven by the chemi-
108
M. Ortner et al. / Biochemical Engineering Journal 99 (2015) 107–113
osmotic gradient across the cell membrane as well as a slow, active and high specific one [9]. Beside uptake as free ions, metals can also be taken up in the form of complexes, such as vitamin B12 [10], cocitrate [11] or bound to siderophores [12]. Trace metals present in the digester undergo complex physic-chemical processes. Their form whether existing as free ions, complex bound or as precipitates depends on various parameters, such as pH, alkalinity, and presence of sulphuric compounds and excreted soluble microbial products (SMPs)[13–15]. To overcome deficits, a broad spectrum of commercial mineral supplementation products are available, designed to address the specific needs for trace elements. On the other hand, addition of trace metals also implies the risk of overdosing, in this way causing toxic effects on the microbial consortium of anaerobic digestion [4]. High heavy metal concentrations may also limit the proper use of digestate as fertilizer and can cause environmental pollution. Adequate supplementation requires appropriate knowledge on the bioavailability of the considered elements. However, on this specific issue little knowledge is available because typical measurement of trace metals present in anaerobic fermenter content usually comprise total concentration and hence does not provide adequate information about its bioavailability and toxicity [16]. In order to get such distribution patterns, sequential extraction techniques need to be applied. The speciation, i.e. the distribution into different fractions (free, adsorbed, precipitated, and unavailable) indicates their availability for metabolic activity [17,18]. Only a few studies investigated trace elements in AD processes using such sequential extraction methods [15,19,20]. The authors demonstrated, that depending on the matrix bioavailable fractions can substantially differ from the total concentration of the trace elements. In the present study slaughterhouse waste with extremely high total nitrogen concentrations (TKN 9 g/kg) was used as the sole substrate in four semi-continuous lab-scale mono-digestion experiments. In earlier investigations it was shown that efficient digestion of this specific substrate is strongly dependent on trace metal supplementation [21]. The purpose of the current work was to study the impact of different levels of trace elements on process performance and to apply a sequential extraction technique aiming to link bioavailability of single elements to process improvement. 2. Materials and methods 2.1. Origin of feedstock and inoculum The feedstock applied in all experiments derived from a large abattoir located in Austria which processes 800,000 heads of pig and more than 50,000 heads of cattle per year. For the continuous fermentation tests, material was taken from the storage tank after thermal hygienisation (70 ◦ C, 1 h). This material serves as feedstock for the on-site operated biogas plant (mesophilic process at 38 ◦ C). The samples were kept at 4 ◦ C until use. The inoculum derived from a previous 6l lab-scale experiment using the same feedstock as described above. To keep the microbial consortium viable after finishing this earlier experiment, feeding was continued with a moderate organic loading rate (OLR) of 0.5 kg VS/m3 d. At start-up of the experiments the inoculum was transferred to the 1 l lab-scale digesters. 2.2. Experimental set-up 2.2.1. Semi-continuous fermentation tests Four semi-continuous fermentation tests (SPU 0–3) were conducted using 1 l reactor vessels with a working volume of 800 ml. Three openings in the vessel (two in the top, one in the bottom)
served as gas outlet and for feeding and sample taking, respectively. To maintain constant operation conditions the reactors were placed in an incubation room with a fixed temperature, 38 ± 0.5 ◦ C. Manual feeding was done once a day seven times per week by substitution of part of the reactor content with fresh substrate. A constant organic loading rate (OLR) of 2.2 kg VS/m3 d was applied over the complete operation period of 4 months. The reactors were continually mixed at 250 rpm. Samples were analysed once a week (every Thursday) for the following parameters: volatile fatty acids (VFAs), chemical oxygen demand (COD), total Kjeldahl nitrogen (TKN), ammonia nitrogen (NH4-N), total solid (TS) and volatile solid (VS). After 33 and 91 days samples were taken for sequential extraction (see Section 2.2.3). Biogas quantity was continuously measured with high precision gas counters (MGC-1 V3, Ritter® , Germany) and recalculated to standard conditions. The gas composition (CH4 , CO2 , H2 S and H2 ) was determined twice per week by GC. For this purpose biogas samples (approx. 15 ml) were taken from the head space of the reactors. Duplicate measurements were made using a GC-TCD gas chromatograph (5890 Hewlett Packard Series II). It consists of a single split-splitless injection port, a HP-Plot Q column and a thermal conductivity detector (TCD). Helium was used as carrier gas. Temperature and pH in the reactor were controlled with a standard multi parameter instrument (Multi 340i, WTW). 2.2.2. Routine analyses TS, VS and COD of substrate and reactor content were analysed according to standard methods DIN DEV 38 414 part 2, DIN DEV 38 414 part 3 and DIN DEV 38409-H41-1, respectively. Samples for TKN analysis were digested in a block-digestion unit K-437 (Büchi® ) connected to a Scrubber B-414 at 420 ◦ C followed by subsequent titration conducted via a distillation unit K-370 with built-in titrator (Büchi® ). The standard method according to VDLUFA was slightly modified using NaOH instead of MgO for pH adjustment. In order to avoid NH3 losses, samples were kept on ice before starting the NH4-N analysis. VFAs were determined after protein removal by Carrez precipitation according to standard method DIN 38 414-19 by HPLC (Agilent® ; column COREGEL 87H, ICE Ion 300; solvent H2 SO4 (0.01 M) using the following setting: flow rate 0.05 ml/min; temperature 65 ◦ C; detector systems: multiple wavelength detector (MWD) and refractive index detector (RID). 2.2.3. Trace elements Trace elements (TE) were added in different amounts to the continuous fermentation tests, termed SPU 1–3. Test SPU 0 served as reference and did not receive supplementation. Three trace element solutions (A–C, corresponding to tests SPU 1–3) with increasing amounts were prepared (Table 1). The initial supplementation rate was 10 l solution per day per l working volume in all the reactors. From day 34 on the daily rate was raised to 50 l/l working volume. The respective volume of TE solution was mixed into the substrate immediately before feeding.
Table 1 Composition of trace element solutions (A–C) applied in the semi-continuous fermentation tests. Element
Solution A [mM]
Solution B [mM]
Solution C [mM]
Ni Co Mo Zn Cu Se
26.3 39.1 4.3 8.8 2.3 0.4
78.8 117.3 12.8 26.3 7.0 1.3
236.4 352.0 38.4 78.8 21.0 4.0
M. Ortner et al. / Biochemical Engineering Journal 99 (2015) 107–113
2.2.4. Sequential extraction (SE) To investigate chemical speciation of trace elements in the feedstock (HYG), in the inoculum as well as in the four continuous fermentation tests (SPU 0–3), sequential extraction based on the method of Ortner et al. (2014b) was applied. The principle of this modified Tessier method is the use of different extraction solvents solubilizing specific fractions of metals. An overview is presented in Fig. 1. Details can be found in Ortner et al. (2014b) [19]. Sequential extraction was done in triplicate. For each analysis fresh material with a TS of around 1 g was used. All centrifugation steps were done with a Beckman GS-6 device at the rotation speed of 3000 rpm (1459 g) using polypropylene centrifuge tubes. The filtration steps were done with cellulose filters (Whatman 595½). Washing steps comprised the re-suspension of the pellet within 25 ml of the applied solvent followed by another centrifugation step. All liquid fractions and pellets obtained from the extraction procedure were stored at 4 ◦ C until further analysis. Quantification of the micro and macro elements was performed with an ICP-OE spectroscope (Jobin Yvon Horiba Ultima). For determination of the total amount of elements in the original fresh sample as well as in the residual fraction microwave digestion was performed with nitric acid in a Milestone UltraClave III device at 240 ◦ C with a final pressure of 100 bar for 20 min. The result of the analytical determination is provided in milligram per kilogram total solid content (mg/kg TS). Moreover, the recovery rate of each element was calculated as the sum of the amounts in all single fractions related to the total content in the fresh sample. 3. Results and discussion 3.1. Characterisation of feedstock and inoculum Chemical parameters including micro and macro elements content of both, inoculum and feedstock were analysed at the start of the experiments. In Table 2 chemical composition is shown, the contents of micro and macro elements are illustrated in Fig. 2. Based on the data from sequential extraction, deficiency in certain essential elements can be assumed. Especially the elements Co, Ni and Mo were not present at all or were only found in low
Freshsample
Fig. 1. Flow chart showing the steps undertaken during sequential extraction (modified Tessier method).
109
Table 2 Characterization of feedstock and inoculum used in the semi-continuous fermentation tests.
Feedstock Inoculum
pH []
TS [%]
VS [%]
NH4-N [g/kg]
TKN [g/kg]
COD [g/kg]
6.38 7.79
11.10 3.13
10.23 2.85
2.12 7.32
8.33 9.05
200.03 33.06
concentrations in the bioavailable fractions (f1, f2, see Fig. 1.). Co was below the detection limit, whereas 0.1 mg/kg TS of Ni was only present in fraction 5 (residual) and 0.27 mg/kg TS of Mo in fraction 4 (organic, sulphides). Due to degradation of the feedstock during the digestion process a mobilization of elements to more bioavailable fractions may take place. Nevertheless, total amounts of Co, Ni or Mo were very low and therefore indicate undersupply. Other essential elements, such as Mn and Zn were found in sufficient amounts and also present to a high extent in the bioavailable fractions (f1 and f2). The Ni, Co and Mo concentrations in the bioavailable fractions are below those reported by Takashima et al. (1990) [22] and Zandvoort et al. (2006) [23], which ranged from 0.001 to 0.12 mg/l although they worked with pure cultures. Even total concentrations of the aforementioned elements in the feedstock are below these suggestions. In contrast to the results from feedstock analyses, all relevant elements were found in the inoculum. Important elements such as Ni, Co and Mo were present to higher extent. Nevertheless, the bioavailable fraction in comparison to total amount was moderate (5–20%). In comparison to the aforementioned suggestions by Takashima et al. and Zandvoort et al., Co and Ni were at the lower range of the recommended concentration, whereas Mo concentration was even too low. In Figs. 2 and 3 data on speciation of 14 different elements are shown. Values below the limit of detection are indicated as 0.00. 3.2. Process performance of tests (SPU 0–3) In a recently published study, Ortner et al. (2014a) demonstrated the feasibility of a stable anaerobic mono-digestion of slaughterhouse waste with good specific methane yields despite extremely high TKN concentrations in the range of 9–11 g/kg. One of the key factors to improve process performance was sufficient supply of essential trace elements. Similar conclusions were drawn by Bayr et al. (2012) [24] and Banks et al. (2012) [25] who added trace elements to an anaerobic digestion process applying a diluted slaughterhouse waste single fraction. Unfortunately, the mentioned studies do not contain detailed analyses of the single trace elements or to what extent they should be supplemented. To get more information, four semi-continuous fermentation tests (SPU 0–3) were operated and monitored for a period of 112 days at constant loading conditions (HRT 50 d, OLR 2.2 kg VS/m3 d). Supplementation of trace elements was lowest at SPU 1, medium at SPU 2 and highest in SPU 3, SPU 0 served as negative control. The composition of the additives (TE solution A–C) mixed to the substrate considered the results gained from sequential extraction. They contained trace elements, which are obviously missing or only present in a minor extent in both, the feedstock and the inoculum. In Fig. 4 the results of the process monitoring are shown. The parameters NH4-N concentration and pH, remained almost constant over the whole operation period and are not illustrated. The concentration of NH4-N in all tests ranged between 7.20 and 7.35 g/kg. The pH at SPU 0 and 1 was around 7.80, whereas pH in SPU 2 and 3 was slightly higher (7.85–7.95). Despite continuous trace element supply only slight differences in process performance were observed until day 34. The VFA level at starting point was high in all tests (about 4500 mg/l) and
110
M. Ortner et al. / Biochemical Engineering Journal 99 (2015) 107–113
Fig. 2. Sequential extraction of elements present in the feedstock (values in table: mg/kg TS).
increased during the first month of operation to a maximum of almost 10,000 mg/l in SPU 0–2 and 7500 mg/l in SPU 3. Herein concentration of propionic acid, an important indicator for process imbalances, increased from about 700 mg/l to 1400 mg/l. Specific methane yields in all tests were relatively low (150–200 Nm3 /t COD). It should be kept in mind that trace element content of the inoculum at the start of the experiment was very low and that supplementation was conducted throughout the feeding period. At the given HRT of 50 d, trace element concentration in the reactor content increases only gradually. Therefore it was assumed that in this first period trace element levels were still too low. Ortner et al. (2014) and Bayr et al. (2012) also demonstrated that anaerobic digestion of slaughterhouse wastes with inadequate trace element
supply causes massive process imbalances in form of high VFA accumulation and low methane yields. In order to accelerate enrichment of trace elements in the process, supplementation rate of TE solution was increased from 10 l/l to 50 l/l from day 34 on. In this second operation period differences in process performance became more obvious. VFA concentration decreased in all reactors except the reference (SPU 0), where it remained at high level (9000–12,000 mg/l). There was a clear trend, showing enhanced and faster reduction of VFA when higher supplementation of trace elements was applied. With respect to the specific methane yield, there was only little difference between SPU 2 and 3 (medium and high supplementation). Both values ranged between 250 and 275 Nm3 CH4 /t COD. In comparison to that methane yield in reactor SPU 1 (low supple-
Fig. 3. Sequential extraction of elements present in the inoculum (values in table: mg/kg TS).
M. Ortner et al. / Biochemical Engineering Journal 99 (2015) 107–113
Fig. 4. Influence of different trace element additions on process performance (methane production, VFA accumulation, propionic acid accumulation, COD removal) in three semi-continuous tests (SPU 0–3) with slaughterhouse waste at 38 ◦ C.
mentation) was lower, between 190 and 220 Nm3 /t COD. In the reference experiment, values of only about 175 Nm3 /t COD were observed. Considering the high ammonia concentration in the reactors, a well-known inhibitor for anaerobic digestion, COD removal rates were quite good and ranged between 70 and 80% for SPU 2 and 3. The reference only achieved about 53%. The relative amount of methane in the biogas ranged between 67 and 69% v/v for all experiments including the reference reactor. To mention, H2 S concentration in the biogas were low (between 300 and 400 ppmv) due to high level of Iron present in the reactors. 3.3. Sequential extraction Samples from all four fermentation tests were taken at day 33 (end of operation period 1) and day 91 and sequential extraction was performed to investigate the essential elements of interest. The recovery rates of Ni, Co, Mo and Zn were between 90 and 110% underlining the validity of the measured values. The applied method delivers five different fractions which correspond to different levels of bioavailability. The occurrence of a certain element in a specific fraction is linked to its chemical nature. In general, Ni, Co and Zn are present as cations. In this form they are able to precipitate with anionic compounds such as sulphides or carbonates. Apart from precipitation, there are indications that carbonates and sulphides are also able to form metal complexes (i.e. [MeHS]+ or [Me(HS)2 ]0 ) and may serve in this form as a source of metal for cell uptake [26,27]. Furthermore, cations are also able to adsorb onto cell or particle surfaces. Consequently these elements can be found in all fractions depending on concentration and physicochemical conditions. In contrast, Mo usually appears as highly
111
soluble anion (molybdate, MoO4 2− ) which does not form any precipitates or adsorb onto surfaces (commonly negative charged). However, under anaerobic conditions Mo (VI) might be reduced to Mo (IV) and precipitation in form of MoS2 is very likely in presence of sulphidic compounds [28]. Fig. 5 illustrates the outcome of the speciation of Ni, Co, Mo and Zn. A comparison of the samples from days 33 and 91 (except the reference) showed a significant increase of both the absolute amount in f1 as well as the relative amount (share of bioavailable fraction within the total amount). Elements present in the fraction f1 were readily bioavailable as they occur as dissolved, complex bound or free ions. Despite different dosage rates the concentration levels of all elements in f1 were relatively similar in all reactors at the time of the 1st extraction (day 33). In contrast, at day 91 when the second analyses of trace element distribution was performed, much more pronounced differences were measured. Compared to the 1st, concentration of Ni was more than 2 times (SPU 1), 6 times (SPU 2) and 17 times (SPU 3) times higher at the 2nd extraction. A similar trend was observed for Co and Mo. A different behaviour was noticed for Zn. Even after 91 days of supplementation no significant increase of the readily bioavailable concentration was observed, except reactor SPU 3 where the highest dosage rate was applied. Besides f1, elements present in fraction f2 (exchangeable; adsorbed elements) are also considered to contribute to bioavailability [29,30]. The aforementioned trend of increase with time was also observed for elements in f2. Concentration of Zn in f2 increased, but not to the same extent as Ni or Co. It seems that Zn had a higher affinity to adsorb on cells or particle surfaces than Ni and Co. The concentration of Mo determined in f2, was very small. This fits to observations of Xu et al. (2006) [31] who demonstrated that adsorption of molybdate to inorganic particles is depending on pH. Significant adsorption occurs only at low pH (<5), whereas at pH 8 only 10–15% of total molybdate is adsorbed. No clear trends was observed in fraction 3 and 4 (carbonate bound and sulphide/organic bound) but generally a significant increase in the 2nd extraction occurred. The final fraction, f5, is considered to be unavailable for cell uptake. Interestingly it was observed that the absolute amounts in f5 remained relatively constant or even decreased for some elements. In the reference reactor (SPU 0) the concentration of all elements decreased with operation time. In particular the share of easily bioavailable elements (f1) was significantly reduced or, in case of Ni and Mo, concentration were even below the detection limit. In f2 concentration of elements remained almost constant or decreased slightly. An exception was Zn. In the 2nd extraction the f2 concentration was more than doubled, probably due to mobilization from other fractions. Beside Ni, Co, Mo and Zn, also Se were analysed. Total amount of Se in the feedstock was 0.8 mg/l TS. During 2nd sequential extraction concentration of about 0.08–0.25 were observed in bioavailable fractions of test SPU 1–3. However, due to the low recovery rates of only 50–70% and the associated analytical uncertainty these values are not further discussed here.
3.4. Comparison of process performance and chemical speciation The results showed an obvious correlation between methane production, VFA accumulation and the share of Ni, Co and Mo in the bioavailable fractions. The higher the concentration of elements observed in f1 or f2, the better the process performance. A distinct concentration of trace elements addition was necessary to have sufficient elements present in bioavailable forms that can be taken up. SPU 2 (medium dosage) and SPU 3 (high dosage) showed almost identical process performance. Both test runs pro-
112
M. Ortner et al. / Biochemical Engineering Journal 99 (2015) 107–113
Fig. 5. Comparison of sequential extraction of Ni, Co, Mo and Zn after 33 and 91 days of operation of tests SPU 0–3. Results are presented in two ways: relative distribution in the different fractions (left) and absolute values (right).
vided similar methane yields and COD removal rates. Only a faster degradation of VFAs accumulated in the first period was observed in SPU 3. Sequential extraction showed in both cases similar shares of Ni, Co and Zn in fractions f1 and f2 (60–70%). The share of Mo was even higher, about 90%. However, in absolute values, total concentrations in f1 and f2 were about 3–5 times higher in SPU 3 compared to SPU 2.
Process performance of test SPU 1 (low dosage) was better than the reference (no dosage) but significantly worse in comparison to SPU 2 and 3. Based on these observations it can be concluded that the optimum trace element supplementation for the given feedstock is in the range of the concentration found in SPU 2. Typically reported stimulatory concentrations for Ni range between 0.05 and 0.6 mg/l fresh matter, for Co between 0.03 and
M. Ortner et al. / Biochemical Engineering Journal 99 (2015) 107–113
10.0 mg/l and for Mo between 0.05 and 50.0 mg/l [6,7,22]. In SPU 1 3.8 mg/l Ni, 7.1 mg/l Co and 0.8 mg/l Mo were analysed (day 91) in the bioavailable fractions (f1 + f2). These concentrations were too low to guarantee optimal performance. On the other hand bioavailable concentrations of 7.1 mg/l Ni, 15.7 mg/l Co and 3.2 mg/l Mo as measured in SPU 2 were sufficient to sustain optimum process performance. Under the given conditions the share of elements in f1 + f2 was between 61 and 68% of the total amount. Accordingly, it is estimated that approximately a third of the supplemented amount is unavailable for the microorganisms, e.g. due to precipitation although it is not possible at this stage to specify the quantities. Additionally, also bioavailability of Se may play an important role. Takashima et al. reported stimulatory concentrations for pure cultures between 79 and 790 g/l. Cho and Park (1995) [32] found 620 g Se/kg as optimal for process stimulation. Although in this study no reliable bioavailable amount can be reported, total amount of Se in tests SPU 1–3 were in the range suggested in the literature. Based on the results the following recommendation is given to avoid process limitations for the given substrate (fresh matter): 2.5 g/t Ni, 3.5 g/t Co, 0.6 g/t Mo and 0.05 g/t Se. 4. Conclusion Slaughterhouse waste as the sole feedstock in anaerobic digestion showed significant process limitations due to deficiency in essential trace elements. Therefore supplementation is a prerequisite for a well-functioning plant. A clear correlation between process performance and concentration of dosed trace elements was shown. Detailed sequential analyses of bioavailability of elements demonstrated that a significant part is not available for microbial uptake. This fact must be borne in mind considering that type and nature of feedstock may have strong impact on accessibility of the elements. According to the obtained results the optimum trace elements supplementation was worked out for the given substrate. Acknowledgements The authors thank Rudolf Großfurtner, who financially supported the project, which was in cooperation with the Austrian Competence Center Bioenergy 2020 and co-funded by the Austrian Research Promotion Agency (FFG). Sincere thanks to Walter Somitsch, Julia Stiel and Gregor Voitl for their valuable support. References [1] F.G. Fermoso, J. Bartacek, S. Jansen, P.N.L. Lens, Metal supplementation to UASB bioreactors: from cell-metal interactions to full-scale application, Sci. Total Environ. 407 (2009) 3652–3667, http://dx.doi.org/10.1016/j.scitotenv.2008.10.043. [2] A.L. Zerkle, C.H. House, S.L. Brantley, Biogeochemical signatures through time as inferred from whole microbial genomes, Am. J. Sci. 305 (2005) 467–502. [3] J.B. Glass, V.J. Orphan, Trace metal requirements for microbial enzymes involved in the production and consumption of methane and nitrous oxide, Front. Microbiol. 3 (2012) 1–20, http://dx.doi.org/10.3389/fmicb.2012.00061. [4] C.M. Plugge, B. Jiang, F. aM. de Bok, C. Tsai, A.J.M. Stams, Effect of tungsten and molybdenum on growth of a syntrophic coculture of Syntrophobacter fumaroxidans and Methanospirillum hungatei, Arch. Microbiol. 191 (2009) 55–61, http://dx.doi.org/10.1007/s00203-008-0428-9. [5] H. Pobeheim, B. Munk, J. Johansson, G.M. Guebitz, Influence of trace elements on methane formation from a synthetic model substrate for maize silage, Bioresour. Technol. 101 (2010) 836–839, http://dx.doi.org/10.1016/j.biortech.2009.08.076. [6] B. Demirel, P. Scherer, Trace element requirements of agricultural biogas digesters during biological conversion of renewable biomass to methane, Biomass Bioenergy 35 (2011) 992–998, http://dx.doi.org/10.1016/j.biombioe.2010.12.022.
113
[7] A. Schattauer, E. Abdoun, P. Weiland, M. Plöchl, M. Heiermann, Abundance of trace elements in demonstration biogas plants, Biosyst. Eng. 108 (2011) 57–65, http://dx.doi.org/10.1016/j.biosystemseng.2010.10.010. [8] U. Ermler, On the mechanism of methyl-coenzyme M reductase, Dalton Trans. (2005) 3451–3458, http://dx.doi.org/10.1039/b506697b. [9] D.H. Nies, Microbial heavy-metal resistance, Appl. Microbiol. Biotechnol. 51 (1999) 730–750, http://dx.doi.org/10.1007/s002530051457. [10] A.D. Ferguson, J. Deisenhofer, Metal import through microbial membranes, Cell 116 (2004) 15–24, http://dx.doi.org/10.1016/S0092-8674(03)1030-1034. [11] B.P. Krom, H. Huttinga, J.B. Warner, J.S. Lolkema, Impact of the Mg(2+)-citrate transporter CitM on heavy metal toxicity in Bacillus subtilis, Arch. Microbiol. 178 (2002) 370–375, http://dx.doi.org/10.1007/s00203-002-0465-8. [12] I. Worms, D.F. Simon, C.S. Hassler, K.J. Wilkinson, Bioavailability of trace metals to aquatic microorganisms: importance of chemical, biological and physical processes on biouptake, Biochimie 88 (2006) 1721–1731, http://dx.doi.org/10.1016/j.biochi.2006.09.008. [13] F.G. Fermoso, J. Bartacek, L. Ching Chung, P. Lens, Supplementation of cobalt to UASB reactors by pulse dosing: CoCl2 versus CoEDTA2-pulses, Biochem. Eng. J. 42 (2008) 111–119. [14] M.B. Osuna, E.D. van Hullebusch, M.H. Zandvoort, J. Iza, P.N.L. Lens, Effect of cobalt sorption on metal fractionation in anaerobic granular sludge, J. Env. Qual. 33 (2004) 1256–1270. [15] E.D. van Hullebusch, S. Utomo, M.H. Zandvoort, P.N.L. Lens, Comparison of three sequential extraction procedures to describe metal fractionation in anaerobic granular sludges, Talanta 65 (2005) 549–558, http://dx.doi.org/10.1016/j.talanta.2004.07.024. [16] J.A. Oleszkiewicz, V.K. Sharma, Stimulation and inhibition of anaerobic processes by heavy metals—A review, Biol. Wastes 31 (1) (1990) 45–67, http://dx.doi.org/10.1016/0269-7483(90)90043-R, ISSN 0269-7483. [17] M.B. Osuna, J. Iza, M. Zandvoort, P.N.L. Lens, Essential metal depletion in an anaerobic reactor, Water Sci. Technol. 48 (2003) 1–8 http://www.ncbi.nlm.nih.gov/pubmed/14640193 [18] A.V. Filgueiras, I. Lavilla, C. Bendicho, Chemical sequential extraction for metal partitioning in environmental solid samples, J. Environ. Monit. 4 (2002) 823–857, http://dx.doi.org/10.1039/b207574c. [19] M. Ortner, L. Rachbauer, W. Somitsch, W. Fuchs, Can bioavailability of trace nutrients be measured in anaerobic digestion? Appl. Energy 126 (2014) 190–198, http://dx.doi.org/10.1016/j.apenergy.2014.03.070. [20] J. Gustavsson, S. Shakeri Yekta, C. Sundberg, A. Karlsson, J. Ejlertsson, U. Skyllberg, et al., Bioavailability of cobalt and nickel during anaerobic digestion of sulfur-rich stillage for biogas formation, Appl. Energy 112 (2013) 473–477, http://dx.doi.org/10.1016/j.apenergy.2013.02.009. [21] M. Ortner, K. Leitzinger, S. Skupien, G. Bochmann, W. Fuchs, Efficient anaerobic mono-digestion of N-rich slaughterhouse waste: influence of ammonia, temperature and trace elements, Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.10.023. [22] R.E. Takashima, M. Speece, Mineral requirements for methan fermentation, Crit. Rev. Environ. Control 19 (1990) 465–479 http://dx.doi.org/10.1080/10643389009388378 [23] M.H. Zandvoort, E.D. van Hullebusch, J. Gieteling, P.N.L. Lens, Granular sludge in full-scale anaerobic bioreactors: trace element content and deficiencies, Enzyme Microb. Technol. 39 (2006) 337–346, http://dx.doi.org/10.1016/j.enzmictec.2006.03.034. [24] S. Bayr, O. Pakarinen, A. Korppoo, S. Liuksia, A. Väisänen, P. Kaparaju, et al., Effect of additives on process stability of mesophilic anaerobic monodigestion of pig slaughterhouse waste, Bioresour. Technol. 120 (2012) 106–113, http://dx.doi.org/10.1016/j.biortech.2012.06.009. [25] C.J. Banks, Y. Zhang, Y. Jiang, S. Heaven, Trace element requirements for stable food waste digestion at elevated ammonia concentrations, Bioresour. Technol. 104 (2012) 127–135, http://dx.doi.org/10.1016/j.biortech.2011.10.068. [26] D. Rickard, G.W. Luther, Metal sulfide complexes and clusters, Rev. Mineral. Geochem. 61 (2006) 421–504. [27] S. Jansen, G. Gonzalez-Gil, H.P. van Leeuwen, The impact of Co and Ni speciation on methanogenesis in sulfidic media—biouptake versus metal dissolution, Enzyme Microb. Technol. 40 (2007) 823–830, http://dx.doi.org/10.1016/j.enzmictec.2006.06.019. [28] R.R. Mendel, Molybdenum: biological activity and metabolism, Dalton Trans. (Cambridge, Engl. 2005) (2005). [29] S.F. Aquino, D.C. Stuckey, Bioavailability and toxicity of metal nutrients during anaerobic digestion, J. Environ. Eng. 133 (2007) 28–35. [30] A. Fuentes, M. Lloréns, J. Sáez, M.I. Aguilar, J.F. Ortuno, V.F. Meseguer, Comparative study of six different sludges by sequential speciation of heavy metals, Bioresour. Technol. 99 (2008) 517–525, http://dx.doi.org/10.1016/j.biortech.2007.01.025. [31] N. Xu, C. Christodoulatos, W. Braida, Adsorption of molybdate and tetrathiomolybdate onto pyrite and goethite: effect of pH and competitive anions, Chemosphere 62 (2006) 1726–1735, http://dx.doi.org/10.1016/j.chemosphere.2005.06.025. [32] J.K. Cho, S.C. Park, Biochemical methane potential and solid state anaerobic digestion of Korean food wastes, Bioresour. Technol. 52 (1995) 245–253.
Contents lists available at ScienceDirect
Biochemical Engineering Journal journal homepage: www.elsevier.com/locate/bej
Bioavailability of essential trace elements and their impact on anaerobic digestion of slaughterhouse waste Markus Ortner a,b,∗ , Michael Rameder a,1 , Lydia Rachbauer b,1 , Günther Bochmann b,2 , Werner Fuchs a,3 a b
University of Natural Resources and Life Sciences, IFA Tulln, Institute for Environmental Biotechnology, Konrad Lorenz-Str. 20, A-3430 Tulln, Austria Bioenergy 2020+, Konrad Lorenz Str. 20, A-3430 Tulln, Austria
a r t i c l e
i n f o
Article history: Received 14 January 2015 Received in revised form 2 March 2015 Accepted 24 March 2015 Available online 25 March 2015 Keywords: Anaerobic processes Anaerobic digestion Trace elements Sequential extraction Chemical speciation Bioavailability Biogas Waste treatment Bioprocess monitoring Slaughterhouse waste
a b s t r a c t Slaughterhouse waste is an energy rich feedstock suitable for anaerobic digestion processes. However, chemical characterization showed a deficiency in essential trace elements which are critical for optimal performance of the process. Hence this study investigated the degree of bioavailability of trace elements in four semi-continuous lab-scale AD tests accepting slaughterhouse waste under mesophilic conditions (38 ◦ C) and a moderate organic loading rate of 2.2 kg/m3 d. Parameters, such as volatile fatty acid (VFA) concentration, COD removal rate and specific methane yield were compared to the results of sequential extraction analysis. The highest methane yield (250–275 Nm3 /t COD), lowest accumulation of VFA (<500 mg/l) and high COD removal rate (75–80%) was obtained when the total concentration of 11.4 mg/l Ni, 25.4 mg/l Co and 4.8 mg/l Mo was present in the reactor, of which 62% of Ni and Co, and 68% of Mo were bioavailable for microbial uptake. Based on these results it can be recommended that a supply of 2.5 g/t Ni, 3.5 g/t Co, 0.6 g/t Mo and 0.05 g/t Se provide optimal conditions for anaerobic digestion of slaughterhouse waste. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Optimal supply of trace elements is a prerequisite for microbial growth and metabolism in anaerobic digestion (AD) processes. Consequently, deficiency causes limitation of the activity of the microbial consortium [1]. This is because, methanogenesis, the final metabolic pathway during AD, involves the participation of various metal-rich enzymes such as carbon monoxide dehydrogenase/acetyl-CoA synthase (Cdh) or methyl coenzyme M reductase (Mcr). These enzymes catalyse key metabolic steps and require sufficient supply of Fe, Ni and Co [2,3]. The exact amount of required trace metals of AD may vary depending on the involved microbial species and their methanogenic pathway, but there are
∗ Corresponding author. Tel.: +43 2272 66280 536. E-mail addresses: [email protected] (M. Ortner), [email protected] (M. Rameder), [email protected] (L. Rachbauer), [email protected] (G. Bochmann), [email protected] (W. Fuchs). 1 Tel.: +43 2272 66280 535. 2 Tel.: +43 2272 66280 536. 3 Tel.: +43 2272 66280 553. http://dx.doi.org/10.1016/j.bej.2015.03.021 1369-703X/© 2015 Elsevier B.V. All rights reserved.
some general tendencies. Fe is the most abundant element, followed by Ni and Co, and smaller amounts of Mo, W and Zn. Fe is part of Fe–S clusters, which are used for both, electron transport and/or catalysis. Ni is either bound to Fe–S clusters or in the active centre of the co-factor F430. Co is present in cobamides involved in methyl group transfer. Mo and W are non-covalent bound to co-factors molybdopterin and tungstopterin, which catalyse two electron redox reactions (reduction of CO2 to formate by formate dehydrogenase (FDH)) [3]. Furthermore, Mo also plays a crucial role in the syntrophic propionate oxidation [4]. In numerous research studies the positive impact of trace element supplementation on anaerobic digestion processes was demonstrated. It was reported that productivity, bio-methanation rate and hence process stability were significantly improved. However, uncertainty about the exact concentration of essential elements that need to be present in a well operating bio-process remains. Suggested values cover a wide spectrum of concentrations comprising a range of more than three decimal powers [5–7]. A major factor of influence is that presence of high metal concentrations does not necessarily imply that microorganisms are also able to take them up and incorporate them into catalytic centres of their enzymes [8]. Microorganisms are using two ways for metal uptake, a fast, passive and unspecific one, driven by the chemi-
108
M. Ortner et al. / Biochemical Engineering Journal 99 (2015) 107–113
osmotic gradient across the cell membrane as well as a slow, active and high specific one [9]. Beside uptake as free ions, metals can also be taken up in the form of complexes, such as vitamin B12 [10], cocitrate [11] or bound to siderophores [12]. Trace metals present in the digester undergo complex physic-chemical processes. Their form whether existing as free ions, complex bound or as precipitates depends on various parameters, such as pH, alkalinity, and presence of sulphuric compounds and excreted soluble microbial products (SMPs)[13–15]. To overcome deficits, a broad spectrum of commercial mineral supplementation products are available, designed to address the specific needs for trace elements. On the other hand, addition of trace metals also implies the risk of overdosing, in this way causing toxic effects on the microbial consortium of anaerobic digestion [4]. High heavy metal concentrations may also limit the proper use of digestate as fertilizer and can cause environmental pollution. Adequate supplementation requires appropriate knowledge on the bioavailability of the considered elements. However, on this specific issue little knowledge is available because typical measurement of trace metals present in anaerobic fermenter content usually comprise total concentration and hence does not provide adequate information about its bioavailability and toxicity [16]. In order to get such distribution patterns, sequential extraction techniques need to be applied. The speciation, i.e. the distribution into different fractions (free, adsorbed, precipitated, and unavailable) indicates their availability for metabolic activity [17,18]. Only a few studies investigated trace elements in AD processes using such sequential extraction methods [15,19,20]. The authors demonstrated, that depending on the matrix bioavailable fractions can substantially differ from the total concentration of the trace elements. In the present study slaughterhouse waste with extremely high total nitrogen concentrations (TKN 9 g/kg) was used as the sole substrate in four semi-continuous lab-scale mono-digestion experiments. In earlier investigations it was shown that efficient digestion of this specific substrate is strongly dependent on trace metal supplementation [21]. The purpose of the current work was to study the impact of different levels of trace elements on process performance and to apply a sequential extraction technique aiming to link bioavailability of single elements to process improvement. 2. Materials and methods 2.1. Origin of feedstock and inoculum The feedstock applied in all experiments derived from a large abattoir located in Austria which processes 800,000 heads of pig and more than 50,000 heads of cattle per year. For the continuous fermentation tests, material was taken from the storage tank after thermal hygienisation (70 ◦ C, 1 h). This material serves as feedstock for the on-site operated biogas plant (mesophilic process at 38 ◦ C). The samples were kept at 4 ◦ C until use. The inoculum derived from a previous 6l lab-scale experiment using the same feedstock as described above. To keep the microbial consortium viable after finishing this earlier experiment, feeding was continued with a moderate organic loading rate (OLR) of 0.5 kg VS/m3 d. At start-up of the experiments the inoculum was transferred to the 1 l lab-scale digesters. 2.2. Experimental set-up 2.2.1. Semi-continuous fermentation tests Four semi-continuous fermentation tests (SPU 0–3) were conducted using 1 l reactor vessels with a working volume of 800 ml. Three openings in the vessel (two in the top, one in the bottom)
served as gas outlet and for feeding and sample taking, respectively. To maintain constant operation conditions the reactors were placed in an incubation room with a fixed temperature, 38 ± 0.5 ◦ C. Manual feeding was done once a day seven times per week by substitution of part of the reactor content with fresh substrate. A constant organic loading rate (OLR) of 2.2 kg VS/m3 d was applied over the complete operation period of 4 months. The reactors were continually mixed at 250 rpm. Samples were analysed once a week (every Thursday) for the following parameters: volatile fatty acids (VFAs), chemical oxygen demand (COD), total Kjeldahl nitrogen (TKN), ammonia nitrogen (NH4-N), total solid (TS) and volatile solid (VS). After 33 and 91 days samples were taken for sequential extraction (see Section 2.2.3). Biogas quantity was continuously measured with high precision gas counters (MGC-1 V3, Ritter® , Germany) and recalculated to standard conditions. The gas composition (CH4 , CO2 , H2 S and H2 ) was determined twice per week by GC. For this purpose biogas samples (approx. 15 ml) were taken from the head space of the reactors. Duplicate measurements were made using a GC-TCD gas chromatograph (5890 Hewlett Packard Series II). It consists of a single split-splitless injection port, a HP-Plot Q column and a thermal conductivity detector (TCD). Helium was used as carrier gas. Temperature and pH in the reactor were controlled with a standard multi parameter instrument (Multi 340i, WTW). 2.2.2. Routine analyses TS, VS and COD of substrate and reactor content were analysed according to standard methods DIN DEV 38 414 part 2, DIN DEV 38 414 part 3 and DIN DEV 38409-H41-1, respectively. Samples for TKN analysis were digested in a block-digestion unit K-437 (Büchi® ) connected to a Scrubber B-414 at 420 ◦ C followed by subsequent titration conducted via a distillation unit K-370 with built-in titrator (Büchi® ). The standard method according to VDLUFA was slightly modified using NaOH instead of MgO for pH adjustment. In order to avoid NH3 losses, samples were kept on ice before starting the NH4-N analysis. VFAs were determined after protein removal by Carrez precipitation according to standard method DIN 38 414-19 by HPLC (Agilent® ; column COREGEL 87H, ICE Ion 300; solvent H2 SO4 (0.01 M) using the following setting: flow rate 0.05 ml/min; temperature 65 ◦ C; detector systems: multiple wavelength detector (MWD) and refractive index detector (RID). 2.2.3. Trace elements Trace elements (TE) were added in different amounts to the continuous fermentation tests, termed SPU 1–3. Test SPU 0 served as reference and did not receive supplementation. Three trace element solutions (A–C, corresponding to tests SPU 1–3) with increasing amounts were prepared (Table 1). The initial supplementation rate was 10 l solution per day per l working volume in all the reactors. From day 34 on the daily rate was raised to 50 l/l working volume. The respective volume of TE solution was mixed into the substrate immediately before feeding.
Table 1 Composition of trace element solutions (A–C) applied in the semi-continuous fermentation tests. Element
Solution A [mM]
Solution B [mM]
Solution C [mM]
Ni Co Mo Zn Cu Se
26.3 39.1 4.3 8.8 2.3 0.4
78.8 117.3 12.8 26.3 7.0 1.3
236.4 352.0 38.4 78.8 21.0 4.0
M. Ortner et al. / Biochemical Engineering Journal 99 (2015) 107–113
2.2.4. Sequential extraction (SE) To investigate chemical speciation of trace elements in the feedstock (HYG), in the inoculum as well as in the four continuous fermentation tests (SPU 0–3), sequential extraction based on the method of Ortner et al. (2014b) was applied. The principle of this modified Tessier method is the use of different extraction solvents solubilizing specific fractions of metals. An overview is presented in Fig. 1. Details can be found in Ortner et al. (2014b) [19]. Sequential extraction was done in triplicate. For each analysis fresh material with a TS of around 1 g was used. All centrifugation steps were done with a Beckman GS-6 device at the rotation speed of 3000 rpm (1459 g) using polypropylene centrifuge tubes. The filtration steps were done with cellulose filters (Whatman 595½). Washing steps comprised the re-suspension of the pellet within 25 ml of the applied solvent followed by another centrifugation step. All liquid fractions and pellets obtained from the extraction procedure were stored at 4 ◦ C until further analysis. Quantification of the micro and macro elements was performed with an ICP-OE spectroscope (Jobin Yvon Horiba Ultima). For determination of the total amount of elements in the original fresh sample as well as in the residual fraction microwave digestion was performed with nitric acid in a Milestone UltraClave III device at 240 ◦ C with a final pressure of 100 bar for 20 min. The result of the analytical determination is provided in milligram per kilogram total solid content (mg/kg TS). Moreover, the recovery rate of each element was calculated as the sum of the amounts in all single fractions related to the total content in the fresh sample. 3. Results and discussion 3.1. Characterisation of feedstock and inoculum Chemical parameters including micro and macro elements content of both, inoculum and feedstock were analysed at the start of the experiments. In Table 2 chemical composition is shown, the contents of micro and macro elements are illustrated in Fig. 2. Based on the data from sequential extraction, deficiency in certain essential elements can be assumed. Especially the elements Co, Ni and Mo were not present at all or were only found in low
Freshsample
Fig. 1. Flow chart showing the steps undertaken during sequential extraction (modified Tessier method).
109
Table 2 Characterization of feedstock and inoculum used in the semi-continuous fermentation tests.
Feedstock Inoculum
pH []
TS [%]
VS [%]
NH4-N [g/kg]
TKN [g/kg]
COD [g/kg]
6.38 7.79
11.10 3.13
10.23 2.85
2.12 7.32
8.33 9.05
200.03 33.06
concentrations in the bioavailable fractions (f1, f2, see Fig. 1.). Co was below the detection limit, whereas 0.1 mg/kg TS of Ni was only present in fraction 5 (residual) and 0.27 mg/kg TS of Mo in fraction 4 (organic, sulphides). Due to degradation of the feedstock during the digestion process a mobilization of elements to more bioavailable fractions may take place. Nevertheless, total amounts of Co, Ni or Mo were very low and therefore indicate undersupply. Other essential elements, such as Mn and Zn were found in sufficient amounts and also present to a high extent in the bioavailable fractions (f1 and f2). The Ni, Co and Mo concentrations in the bioavailable fractions are below those reported by Takashima et al. (1990) [22] and Zandvoort et al. (2006) [23], which ranged from 0.001 to 0.12 mg/l although they worked with pure cultures. Even total concentrations of the aforementioned elements in the feedstock are below these suggestions. In contrast to the results from feedstock analyses, all relevant elements were found in the inoculum. Important elements such as Ni, Co and Mo were present to higher extent. Nevertheless, the bioavailable fraction in comparison to total amount was moderate (5–20%). In comparison to the aforementioned suggestions by Takashima et al. and Zandvoort et al., Co and Ni were at the lower range of the recommended concentration, whereas Mo concentration was even too low. In Figs. 2 and 3 data on speciation of 14 different elements are shown. Values below the limit of detection are indicated as 0.00. 3.2. Process performance of tests (SPU 0–3) In a recently published study, Ortner et al. (2014a) demonstrated the feasibility of a stable anaerobic mono-digestion of slaughterhouse waste with good specific methane yields despite extremely high TKN concentrations in the range of 9–11 g/kg. One of the key factors to improve process performance was sufficient supply of essential trace elements. Similar conclusions were drawn by Bayr et al. (2012) [24] and Banks et al. (2012) [25] who added trace elements to an anaerobic digestion process applying a diluted slaughterhouse waste single fraction. Unfortunately, the mentioned studies do not contain detailed analyses of the single trace elements or to what extent they should be supplemented. To get more information, four semi-continuous fermentation tests (SPU 0–3) were operated and monitored for a period of 112 days at constant loading conditions (HRT 50 d, OLR 2.2 kg VS/m3 d). Supplementation of trace elements was lowest at SPU 1, medium at SPU 2 and highest in SPU 3, SPU 0 served as negative control. The composition of the additives (TE solution A–C) mixed to the substrate considered the results gained from sequential extraction. They contained trace elements, which are obviously missing or only present in a minor extent in both, the feedstock and the inoculum. In Fig. 4 the results of the process monitoring are shown. The parameters NH4-N concentration and pH, remained almost constant over the whole operation period and are not illustrated. The concentration of NH4-N in all tests ranged between 7.20 and 7.35 g/kg. The pH at SPU 0 and 1 was around 7.80, whereas pH in SPU 2 and 3 was slightly higher (7.85–7.95). Despite continuous trace element supply only slight differences in process performance were observed until day 34. The VFA level at starting point was high in all tests (about 4500 mg/l) and
110
M. Ortner et al. / Biochemical Engineering Journal 99 (2015) 107–113
Fig. 2. Sequential extraction of elements present in the feedstock (values in table: mg/kg TS).
increased during the first month of operation to a maximum of almost 10,000 mg/l in SPU 0–2 and 7500 mg/l in SPU 3. Herein concentration of propionic acid, an important indicator for process imbalances, increased from about 700 mg/l to 1400 mg/l. Specific methane yields in all tests were relatively low (150–200 Nm3 /t COD). It should be kept in mind that trace element content of the inoculum at the start of the experiment was very low and that supplementation was conducted throughout the feeding period. At the given HRT of 50 d, trace element concentration in the reactor content increases only gradually. Therefore it was assumed that in this first period trace element levels were still too low. Ortner et al. (2014) and Bayr et al. (2012) also demonstrated that anaerobic digestion of slaughterhouse wastes with inadequate trace element
supply causes massive process imbalances in form of high VFA accumulation and low methane yields. In order to accelerate enrichment of trace elements in the process, supplementation rate of TE solution was increased from 10 l/l to 50 l/l from day 34 on. In this second operation period differences in process performance became more obvious. VFA concentration decreased in all reactors except the reference (SPU 0), where it remained at high level (9000–12,000 mg/l). There was a clear trend, showing enhanced and faster reduction of VFA when higher supplementation of trace elements was applied. With respect to the specific methane yield, there was only little difference between SPU 2 and 3 (medium and high supplementation). Both values ranged between 250 and 275 Nm3 CH4 /t COD. In comparison to that methane yield in reactor SPU 1 (low supple-
Fig. 3. Sequential extraction of elements present in the inoculum (values in table: mg/kg TS).
M. Ortner et al. / Biochemical Engineering Journal 99 (2015) 107–113
Fig. 4. Influence of different trace element additions on process performance (methane production, VFA accumulation, propionic acid accumulation, COD removal) in three semi-continuous tests (SPU 0–3) with slaughterhouse waste at 38 ◦ C.
mentation) was lower, between 190 and 220 Nm3 /t COD. In the reference experiment, values of only about 175 Nm3 /t COD were observed. Considering the high ammonia concentration in the reactors, a well-known inhibitor for anaerobic digestion, COD removal rates were quite good and ranged between 70 and 80% for SPU 2 and 3. The reference only achieved about 53%. The relative amount of methane in the biogas ranged between 67 and 69% v/v for all experiments including the reference reactor. To mention, H2 S concentration in the biogas were low (between 300 and 400 ppmv) due to high level of Iron present in the reactors. 3.3. Sequential extraction Samples from all four fermentation tests were taken at day 33 (end of operation period 1) and day 91 and sequential extraction was performed to investigate the essential elements of interest. The recovery rates of Ni, Co, Mo and Zn were between 90 and 110% underlining the validity of the measured values. The applied method delivers five different fractions which correspond to different levels of bioavailability. The occurrence of a certain element in a specific fraction is linked to its chemical nature. In general, Ni, Co and Zn are present as cations. In this form they are able to precipitate with anionic compounds such as sulphides or carbonates. Apart from precipitation, there are indications that carbonates and sulphides are also able to form metal complexes (i.e. [MeHS]+ or [Me(HS)2 ]0 ) and may serve in this form as a source of metal for cell uptake [26,27]. Furthermore, cations are also able to adsorb onto cell or particle surfaces. Consequently these elements can be found in all fractions depending on concentration and physicochemical conditions. In contrast, Mo usually appears as highly
111
soluble anion (molybdate, MoO4 2− ) which does not form any precipitates or adsorb onto surfaces (commonly negative charged). However, under anaerobic conditions Mo (VI) might be reduced to Mo (IV) and precipitation in form of MoS2 is very likely in presence of sulphidic compounds [28]. Fig. 5 illustrates the outcome of the speciation of Ni, Co, Mo and Zn. A comparison of the samples from days 33 and 91 (except the reference) showed a significant increase of both the absolute amount in f1 as well as the relative amount (share of bioavailable fraction within the total amount). Elements present in the fraction f1 were readily bioavailable as they occur as dissolved, complex bound or free ions. Despite different dosage rates the concentration levels of all elements in f1 were relatively similar in all reactors at the time of the 1st extraction (day 33). In contrast, at day 91 when the second analyses of trace element distribution was performed, much more pronounced differences were measured. Compared to the 1st, concentration of Ni was more than 2 times (SPU 1), 6 times (SPU 2) and 17 times (SPU 3) times higher at the 2nd extraction. A similar trend was observed for Co and Mo. A different behaviour was noticed for Zn. Even after 91 days of supplementation no significant increase of the readily bioavailable concentration was observed, except reactor SPU 3 where the highest dosage rate was applied. Besides f1, elements present in fraction f2 (exchangeable; adsorbed elements) are also considered to contribute to bioavailability [29,30]. The aforementioned trend of increase with time was also observed for elements in f2. Concentration of Zn in f2 increased, but not to the same extent as Ni or Co. It seems that Zn had a higher affinity to adsorb on cells or particle surfaces than Ni and Co. The concentration of Mo determined in f2, was very small. This fits to observations of Xu et al. (2006) [31] who demonstrated that adsorption of molybdate to inorganic particles is depending on pH. Significant adsorption occurs only at low pH (<5), whereas at pH 8 only 10–15% of total molybdate is adsorbed. No clear trends was observed in fraction 3 and 4 (carbonate bound and sulphide/organic bound) but generally a significant increase in the 2nd extraction occurred. The final fraction, f5, is considered to be unavailable for cell uptake. Interestingly it was observed that the absolute amounts in f5 remained relatively constant or even decreased for some elements. In the reference reactor (SPU 0) the concentration of all elements decreased with operation time. In particular the share of easily bioavailable elements (f1) was significantly reduced or, in case of Ni and Mo, concentration were even below the detection limit. In f2 concentration of elements remained almost constant or decreased slightly. An exception was Zn. In the 2nd extraction the f2 concentration was more than doubled, probably due to mobilization from other fractions. Beside Ni, Co, Mo and Zn, also Se were analysed. Total amount of Se in the feedstock was 0.8 mg/l TS. During 2nd sequential extraction concentration of about 0.08–0.25 were observed in bioavailable fractions of test SPU 1–3. However, due to the low recovery rates of only 50–70% and the associated analytical uncertainty these values are not further discussed here.
3.4. Comparison of process performance and chemical speciation The results showed an obvious correlation between methane production, VFA accumulation and the share of Ni, Co and Mo in the bioavailable fractions. The higher the concentration of elements observed in f1 or f2, the better the process performance. A distinct concentration of trace elements addition was necessary to have sufficient elements present in bioavailable forms that can be taken up. SPU 2 (medium dosage) and SPU 3 (high dosage) showed almost identical process performance. Both test runs pro-
112
M. Ortner et al. / Biochemical Engineering Journal 99 (2015) 107–113
Fig. 5. Comparison of sequential extraction of Ni, Co, Mo and Zn after 33 and 91 days of operation of tests SPU 0–3. Results are presented in two ways: relative distribution in the different fractions (left) and absolute values (right).
vided similar methane yields and COD removal rates. Only a faster degradation of VFAs accumulated in the first period was observed in SPU 3. Sequential extraction showed in both cases similar shares of Ni, Co and Zn in fractions f1 and f2 (60–70%). The share of Mo was even higher, about 90%. However, in absolute values, total concentrations in f1 and f2 were about 3–5 times higher in SPU 3 compared to SPU 2.
Process performance of test SPU 1 (low dosage) was better than the reference (no dosage) but significantly worse in comparison to SPU 2 and 3. Based on these observations it can be concluded that the optimum trace element supplementation for the given feedstock is in the range of the concentration found in SPU 2. Typically reported stimulatory concentrations for Ni range between 0.05 and 0.6 mg/l fresh matter, for Co between 0.03 and
M. Ortner et al. / Biochemical Engineering Journal 99 (2015) 107–113
10.0 mg/l and for Mo between 0.05 and 50.0 mg/l [6,7,22]. In SPU 1 3.8 mg/l Ni, 7.1 mg/l Co and 0.8 mg/l Mo were analysed (day 91) in the bioavailable fractions (f1 + f2). These concentrations were too low to guarantee optimal performance. On the other hand bioavailable concentrations of 7.1 mg/l Ni, 15.7 mg/l Co and 3.2 mg/l Mo as measured in SPU 2 were sufficient to sustain optimum process performance. Under the given conditions the share of elements in f1 + f2 was between 61 and 68% of the total amount. Accordingly, it is estimated that approximately a third of the supplemented amount is unavailable for the microorganisms, e.g. due to precipitation although it is not possible at this stage to specify the quantities. Additionally, also bioavailability of Se may play an important role. Takashima et al. reported stimulatory concentrations for pure cultures between 79 and 790 g/l. Cho and Park (1995) [32] found 620 g Se/kg as optimal for process stimulation. Although in this study no reliable bioavailable amount can be reported, total amount of Se in tests SPU 1–3 were in the range suggested in the literature. Based on the results the following recommendation is given to avoid process limitations for the given substrate (fresh matter): 2.5 g/t Ni, 3.5 g/t Co, 0.6 g/t Mo and 0.05 g/t Se. 4. Conclusion Slaughterhouse waste as the sole feedstock in anaerobic digestion showed significant process limitations due to deficiency in essential trace elements. Therefore supplementation is a prerequisite for a well-functioning plant. A clear correlation between process performance and concentration of dosed trace elements was shown. Detailed sequential analyses of bioavailability of elements demonstrated that a significant part is not available for microbial uptake. This fact must be borne in mind considering that type and nature of feedstock may have strong impact on accessibility of the elements. According to the obtained results the optimum trace elements supplementation was worked out for the given substrate. Acknowledgements The authors thank Rudolf Großfurtner, who financially supported the project, which was in cooperation with the Austrian Competence Center Bioenergy 2020 and co-funded by the Austrian Research Promotion Agency (FFG). Sincere thanks to Walter Somitsch, Julia Stiel and Gregor Voitl for their valuable support. References [1] F.G. Fermoso, J. Bartacek, S. Jansen, P.N.L. Lens, Metal supplementation to UASB bioreactors: from cell-metal interactions to full-scale application, Sci. Total Environ. 407 (2009) 3652–3667, http://dx.doi.org/10.1016/j.scitotenv.2008.10.043. [2] A.L. Zerkle, C.H. House, S.L. Brantley, Biogeochemical signatures through time as inferred from whole microbial genomes, Am. J. Sci. 305 (2005) 467–502. [3] J.B. Glass, V.J. Orphan, Trace metal requirements for microbial enzymes involved in the production and consumption of methane and nitrous oxide, Front. Microbiol. 3 (2012) 1–20, http://dx.doi.org/10.3389/fmicb.2012.00061. [4] C.M. Plugge, B. Jiang, F. aM. de Bok, C. Tsai, A.J.M. Stams, Effect of tungsten and molybdenum on growth of a syntrophic coculture of Syntrophobacter fumaroxidans and Methanospirillum hungatei, Arch. Microbiol. 191 (2009) 55–61, http://dx.doi.org/10.1007/s00203-008-0428-9. [5] H. Pobeheim, B. Munk, J. Johansson, G.M. Guebitz, Influence of trace elements on methane formation from a synthetic model substrate for maize silage, Bioresour. Technol. 101 (2010) 836–839, http://dx.doi.org/10.1016/j.biortech.2009.08.076. [6] B. Demirel, P. Scherer, Trace element requirements of agricultural biogas digesters during biological conversion of renewable biomass to methane, Biomass Bioenergy 35 (2011) 992–998, http://dx.doi.org/10.1016/j.biombioe.2010.12.022.
113
[7] A. Schattauer, E. Abdoun, P. Weiland, M. Plöchl, M. Heiermann, Abundance of trace elements in demonstration biogas plants, Biosyst. Eng. 108 (2011) 57–65, http://dx.doi.org/10.1016/j.biosystemseng.2010.10.010. [8] U. Ermler, On the mechanism of methyl-coenzyme M reductase, Dalton Trans. (2005) 3451–3458, http://dx.doi.org/10.1039/b506697b. [9] D.H. Nies, Microbial heavy-metal resistance, Appl. Microbiol. Biotechnol. 51 (1999) 730–750, http://dx.doi.org/10.1007/s002530051457. [10] A.D. Ferguson, J. Deisenhofer, Metal import through microbial membranes, Cell 116 (2004) 15–24, http://dx.doi.org/10.1016/S0092-8674(03)1030-1034. [11] B.P. Krom, H. Huttinga, J.B. Warner, J.S. Lolkema, Impact of the Mg(2+)-citrate transporter CitM on heavy metal toxicity in Bacillus subtilis, Arch. Microbiol. 178 (2002) 370–375, http://dx.doi.org/10.1007/s00203-002-0465-8. [12] I. Worms, D.F. Simon, C.S. Hassler, K.J. Wilkinson, Bioavailability of trace metals to aquatic microorganisms: importance of chemical, biological and physical processes on biouptake, Biochimie 88 (2006) 1721–1731, http://dx.doi.org/10.1016/j.biochi.2006.09.008. [13] F.G. Fermoso, J. Bartacek, L. Ching Chung, P. Lens, Supplementation of cobalt to UASB reactors by pulse dosing: CoCl2 versus CoEDTA2-pulses, Biochem. Eng. J. 42 (2008) 111–119. [14] M.B. Osuna, E.D. van Hullebusch, M.H. Zandvoort, J. Iza, P.N.L. Lens, Effect of cobalt sorption on metal fractionation in anaerobic granular sludge, J. Env. Qual. 33 (2004) 1256–1270. [15] E.D. van Hullebusch, S. Utomo, M.H. Zandvoort, P.N.L. Lens, Comparison of three sequential extraction procedures to describe metal fractionation in anaerobic granular sludges, Talanta 65 (2005) 549–558, http://dx.doi.org/10.1016/j.talanta.2004.07.024. [16] J.A. Oleszkiewicz, V.K. Sharma, Stimulation and inhibition of anaerobic processes by heavy metals—A review, Biol. Wastes 31 (1) (1990) 45–67, http://dx.doi.org/10.1016/0269-7483(90)90043-R, ISSN 0269-7483. [17] M.B. Osuna, J. Iza, M. Zandvoort, P.N.L. Lens, Essential metal depletion in an anaerobic reactor, Water Sci. Technol. 48 (2003) 1–8 http://www.ncbi.nlm.nih.gov/pubmed/14640193 [18] A.V. Filgueiras, I. Lavilla, C. Bendicho, Chemical sequential extraction for metal partitioning in environmental solid samples, J. Environ. Monit. 4 (2002) 823–857, http://dx.doi.org/10.1039/b207574c. [19] M. Ortner, L. Rachbauer, W. Somitsch, W. Fuchs, Can bioavailability of trace nutrients be measured in anaerobic digestion? Appl. Energy 126 (2014) 190–198, http://dx.doi.org/10.1016/j.apenergy.2014.03.070. [20] J. Gustavsson, S. Shakeri Yekta, C. Sundberg, A. Karlsson, J. Ejlertsson, U. Skyllberg, et al., Bioavailability of cobalt and nickel during anaerobic digestion of sulfur-rich stillage for biogas formation, Appl. Energy 112 (2013) 473–477, http://dx.doi.org/10.1016/j.apenergy.2013.02.009. [21] M. Ortner, K. Leitzinger, S. Skupien, G. Bochmann, W. Fuchs, Efficient anaerobic mono-digestion of N-rich slaughterhouse waste: influence of ammonia, temperature and trace elements, Bioresour. Technol. (2014), http://dx.doi.org/10.1016/j.biortech.2014.10.023. [22] R.E. Takashima, M. Speece, Mineral requirements for methan fermentation, Crit. Rev. Environ. Control 19 (1990) 465–479 http://dx.doi.org/10.1080/10643389009388378 [23] M.H. Zandvoort, E.D. van Hullebusch, J. Gieteling, P.N.L. Lens, Granular sludge in full-scale anaerobic bioreactors: trace element content and deficiencies, Enzyme Microb. Technol. 39 (2006) 337–346, http://dx.doi.org/10.1016/j.enzmictec.2006.03.034. [24] S. Bayr, O. Pakarinen, A. Korppoo, S. Liuksia, A. Väisänen, P. Kaparaju, et al., Effect of additives on process stability of mesophilic anaerobic monodigestion of pig slaughterhouse waste, Bioresour. Technol. 120 (2012) 106–113, http://dx.doi.org/10.1016/j.biortech.2012.06.009. [25] C.J. Banks, Y. Zhang, Y. Jiang, S. Heaven, Trace element requirements for stable food waste digestion at elevated ammonia concentrations, Bioresour. Technol. 104 (2012) 127–135, http://dx.doi.org/10.1016/j.biortech.2011.10.068. [26] D. Rickard, G.W. Luther, Metal sulfide complexes and clusters, Rev. Mineral. Geochem. 61 (2006) 421–504. [27] S. Jansen, G. Gonzalez-Gil, H.P. van Leeuwen, The impact of Co and Ni speciation on methanogenesis in sulfidic media—biouptake versus metal dissolution, Enzyme Microb. Technol. 40 (2007) 823–830, http://dx.doi.org/10.1016/j.enzmictec.2006.06.019. [28] R.R. Mendel, Molybdenum: biological activity and metabolism, Dalton Trans. (Cambridge, Engl. 2005) (2005). [29] S.F. Aquino, D.C. Stuckey, Bioavailability and toxicity of metal nutrients during anaerobic digestion, J. Environ. Eng. 133 (2007) 28–35. [30] A. Fuentes, M. Lloréns, J. Sáez, M.I. Aguilar, J.F. Ortuno, V.F. Meseguer, Comparative study of six different sludges by sequential speciation of heavy metals, Bioresour. Technol. 99 (2008) 517–525, http://dx.doi.org/10.1016/j.biortech.2007.01.025. [31] N. Xu, C. Christodoulatos, W. Braida, Adsorption of molybdate and tetrathiomolybdate onto pyrite and goethite: effect of pH and competitive anions, Chemosphere 62 (2006) 1726–1735, http://dx.doi.org/10.1016/j.chemosphere.2005.06.025. [32] J.K. Cho, S.C. Park, Biochemical methane potential and solid state anaerobic digestion of Korean food wastes, Bioresour. Technol. 52 (1995) 245–253.