Abstract
The anaerobic mesophilic co-digestion of sugar beet pulp stillage with poultry manure and municipal sewage sludge was investigated in this study. The sugar beet pulp stillage (SBPS) mono-digestion failed owing to an accumulation of volatile fatty acids, leading to a pH value lower than 5.5. A 20% addition of poultry manure to stillage allowed for stable digestion performance despite high volatile fatty acid (total volatile fatty acids) concentrations of 5500–8500 g m−3 with propionic acid being the predominant one and constituting 72%–76% total volatile fatty acids. For this mixture, the maximum methane production of 418 dm3 kgVSfed−1 was achieved when the reactor was operated at a solids retention time of 20 days and an organic loading rate of 4.25 kgVS m−3 d−1. The co-digestion of stillage with 60% municipal sewage sludge gave the average methane yield of around 357 dm3 kgVSfed−1 for all operational conditions applied, however, the methane percentage of biogas (up to 70%) was far greater than the corresponding values obtained for sugar beet pulp stillage–poultry manure co-digestion. Neither ammonia nor volatile fatty acids destabilised the biogas production, and the volatile fatty acid profile showed the dominance of acetic acid (72%–82% total volatile fatty acids) followed by propionic and butyric acids.
Introduction
Stillage is a liquid residue obtained during ethanol fermentation of carbohydrate-rich materials. It is estimated that during fermentation, up to 20 L of stillage can be generated for each litre of ethanol produced (Eskicioglu et al., 2011; Kaparaju et al., 2010; Sharma et al., 2013). Because of its relatively high nutritional value, today stillage is mainly used for animal feed. In order to reduce transportation costs and extend durability, stillage is processed through centrifugation and drying before being used as fodder. However, these operations are energy-intensive and may account for up to 45% of the total energy demand at the bioethanol plant (Andalib et al., 2014; Eskicioglu et al., 2011; Westerholm et al., 2012). Moreover, stillage from nonfood sources (e.g. sugarcane bagasse or sugar beet pulp (SBP)) has generally poorer nutritive quality, which restricts its use as animal feed (Tian et al., 2013). An alternative option for stillage utilisation is its use for biogas production. Anaerobic digestion of stillage allows recovery of part of the energy at a bioethanol production plant, and the digestate can be potentially used as fertiliser (Sharma et al., 2013). Anaerobic digestion was applied to stillage of various origins, including corn (Andalib et al., 2014; Eskicioglu et al., 2011), grains (Moestedt et al., 2013), cereals (Westerholm et al., 2012), sugarcane bagasse (Tian et al., 2013), wheat straw (Kaparaju et al., 2010) and kitchen waste (Sun et al., 2014), whereas little has been written about the anaerobic treatment of stillage derived from sugar beet or SBP (Morares et al., 2015). Although relatively high methane yields of up 600 m3 kgVSfed−1 can be achieved from stillage (Alkan-Ozkaynak and Karthikeyan, 2011; Andalib et al., 2014), some researchers have also suggested the need to use a long solids retention time (SRT) of 40–60 d to sustain the growth of acetate-oxidising bacteria as well as methanogens in the digester. Additionally, low buffering capacity combined with low content of some trace elements results in digestion instability, especially when stillage of cellulosic origin is treated (Moestedt et al., 2013; Tian et al., 2013; Westerholm et al., 2012). In light of the above, an attractive solution may be co-digestion of stillage with other organic wastes. In our previous studies (Borowski and Kucner, 2015; Borowski et al., 2016), municipal sewage sludge (MSS) as well as poultry manure (PM) were successfully applied to improve the anaerobic digestion of dewatered SBP after enzymatic hydrolysis. Moreover, batch tests with stillage obtained during bioethanol production from this pulp showed a high potential of this material for both methane and hydrogen production (Berłowska et al., 2017). Hence, this study investigates the semi-continuous anaerobic digestion of sugar beet pulp stillage (SBPS) as the sole substrate, as well as the co-digestion of this material with sewage sludge and poultry manure. Literature concerning the use of animal manure for co-digestion with stillage is scarce (Sharma et al., 2013; Westerholm et al., 2012), whereas, to the best of the authors’ knowledge, there are no studies on the anaerobic treatment of SBPS with sewage sludge.
Materials and methods
Materials
SBPS constituted a by-product of bioethanol fermentation using SBP as a raw material. Details of bioethanol production and SBP processing have been described by Berłowska et al. (2017). In brief, fresh SBP, collected from the Dobrzelin Sugar Factory (Poland), was saccharified for 16 h at 50 °C using a mixture (1:1) of two commercial enzyme preparations: Viscozyme and UltrafloMax (Bagsvaerd, Denmark). Then, the hydrolyzate obtained from saccharification was subjected to ethanol fermentation using mixed cultures of yeasts: Sacharomyces cerevisiae Ethanol Red, Kluyveromyces marxianus NCYC179 and Scheffersomyces stipitis LOCK0047. About 80 dm3 of the resulting stillage was then frozen at −30 °C prior to use.
MSS, collected from the Wastewater Treatment Plant in Łódź, was a mixture of primary and waste activated sludge in the average volume proportion of 1:1. Poultry manure originated from a poultry farm in Zgierz breeding around 28,000 laying hens. Both sewage sludge and poultry manure were stored at 2 °C before use.
The composition of all the substrates is shown in Table 1.
Characteristics of substrates used for the experiments.
SBPS: sugar beet pulp stillage; PM: poultry manure; MSS: municipal sewage sludge; TS: total solids; VS: volatile solids.
Experiments
The experiments were performed in two cylindrical reactors named R1 and R2, each with a total capacity of 5 dm3 and a working volume of 3 dm3. Each reactor was coupled with a 4 dm3 gas collecting tank to provide strict anaerobic conditions and to measure the biogas yield. The reactors were fed and drawn-off once a day with a peristaltic pump. The digestate from the previous experiments with similar substrates (Borowski and Kucner, 2015; Borowski et al., 2016) was used as inoculum, therefore no special acclimation of the digesters was required. The first reactor was initially fed with stillage, which was then supplemented with poultry manure. The second reactor was operated with the mixture of stillage and sewage sludge. The experiments were performed under mesophilic conditions (35 °C). Details of the reactor performance are summarised in Table 2.
Operating parameters of the digesters.
SBPS: sugar beet pulp stillage; PM: poultry manure; MSS: municipal sewage sludge; TS: total solids; VS: volatile solids.
Analyses
Total and volatile solids (TS, VS), pH, and total alkalinity (TAL) were analysed according to standard methods (APHA, 2005). The total ammonium nitrogen, and orthophosphates (PO4−3) were determined using a DR2800 spectrophotometer with HACH-Lange tests no. 8038 and 8048, respectively. The concentration of free ammonia was then calculated according to the formula described by Hansen et al. (1998). Total volatile fatty acids (TVFA) were measured spectrophotometrically with DR2800 and a HACH-Lange method no. LCK365. Individual organic acids were also quantified with a high-performance liquid chromatography as described previously (Borowski and Kucner, 2015). Elemental analysis (carbon, hydrogen, nitrogen and phosphorus) of raw materials was performed with a NA 2500 analyser (CE Instruments, Wigan, UK) following the manufacturer’s procedure. The biogas yield was monitored by a water displacement method as described elsewhere (Cuetos et al., 2011; Sterling et al., 2001). The biogas methane content was measured with a gas analyser, model GA-21 plus (Madur, Poland).
The analyses of individual samples were performed in at least triplicates. The calculations of averages, standard deviations and the analysis of variance (ANOVA) were performed in Microsoft Excel 2010. A confidence level of 0.05 was selected for all statistical comparisons.
Results and discussion
Characteristics of substrates
The composition of substrates used in this study are depicted in Table 1.
SBPS was characterised by a high concentration of both organic solids and nutrients. It had a moisture content of approximately 4.5% with volatile solids comprising of around 83% TS. The average concentrations of nitrogen and phosphorus in SBPS (5.8% TS and 4.92% TS, respectively) were greater than the corresponding data reported in the literature (Andalib et al., 2014; Eskicioglu et al., 2011; Morares et al., 2015). Consequently, the high methane yield of 445 dm3 kgVS−1 was achieved in batch tests performed with this substrate (Berłowska et al., 2017). In contrast, poultry manure displayed the lowest organic and nutrient contents, especially phosphorus, which was also reflected by a relatively low methane yield of 300 dm3 kgVS−1 (Borowski et al., 2016). However, poultry manure had the most favourable carbon to nitrogen (C/N) balance of 12.3, compared with SBPS and sewage sludge (10.7 and 8.7, respectively).
Process performance
The operating parameters and performances of the experiments are summarised in Tables 2 and 3, whereas the profiles of biogas and methane yields are plotted in Figures 1 and 2.
Performances of the digesters.
GPR: gas production rate; SGP: specific biogas production; SMP: specific methane production; VS: volatile solids

Weekly average biogas and methane yields reported during the anaerobic mono-digestion of stillage (R1-stage 1) and co-digestion of stillage with poultry manure (R1-stages 2–4).

Weekly average biogas and methane yields reported during the anaerobic co-digestion of stillage with municipal sewage sludge (R2-stages 1–2).
In the first experimental run (R1-stage 1), the digester was exclusively fed with stillage and operated at an organic loading rate (OLR) of 2.19 kgVS m−3 d−1. Within 2 weeks of operation, the methane yield rose to around 200 dm3 kgVSfed−1, and that value was maintained for another week. At the same time, a rapid growth of volatile fatty acids together with a pH drop were observed (Figure 3). From the 15th day of operation, the biogas and methane productions started decreasing, whereas TVFA reached the value of 17,000 g m−3 at the end of the 6th week. To reverse this negative trend, stillage was mixed with 10% of poultry manure (by weight) and delivered to the digester from the beginning of the 7th week. However, the following 5 weeks of the operation with this mixture did not bring a recovery of the digester. The biogas production ceased and the pH value remained below 5.5. Hence, prior to the next run, the digester was inoculated with a 1 dm3 of anaerobic sludge derived from the wastewater treatment plant in Łódź. The third experimental run (R1-stage 3) was performed with the mixture of stillage and 20% poultry manure. The biogas and methane quickly rose within 2 weeks of operation and pH was established at an average value of 7.5. The digester was successfully operated with this mixture through stages 3 and 4 of R1, at corresponding OLR values of 4.25 and 5.24 kgVS m−3 d−1. The average methane production reported in R1-stage 3 was 418 dm−3 kgVSfed−1, which was the maximum obtained in the study and was also associated with the highest volatile solids reduction of 75%. In stage 4 of R1, the methane yield slightly dropped to 379 dm−3 kgVSfed−1, whereas the VS-reduction rate decreased to 53%, but the digester was operated at OLR considerably greater than a recommended maximum value of 3.5 kgVS m−3 d−1 for wet anaerobic digestion (Weiland, 2010). Moreover, the lower methane yield and volatile solids-reduction rate can be attributed to the higher TVFA production with propionic acid being the predominant one, which is discussed in the next section. The literature concerning the use of bioethanol stillage for co-digestion with poultry manure and other manure types is scarce. Sharma et al. (2013) investigated thermophilic anaerobic digestion of poultry litter with thin stillage. They obtained the highest biogas production of approximately 500 dm3 m−3 d−1 with 67%–69% of methane from the mixture containing 60% of stillage. The authors have not given the specific biogas and methane yields in their study. Morares et al. (2015) conducted the co-digestion experiments with the mixture of vinasse and cow manure (97:3). The highest methane yield obtained in that study was 323 dm3 kgVSfed−1. The co-digestion of cattle manure and stillage was also investigated by Westerholm et al. (2012). The authors reported the increase in methane yield up to 310 dm3 kgVSfed−1 when stillage was supplemented with 15% manure (volatile solids basis).

Changes in pH, volatile fatty acids and ammonium nitrogen concentrations during the anaerobic mono-digestion of stillage (R1-stage 1) and co-digestion of stillage with poultry manure (R1-stages 2-4).
In the R2 runs, the mixture of stillage and sewage sludge used in the experiments was blended at a ratio of 40:60, and the digester was operated at an OLR of 2.14–3.17 kgVS m−3 d−1. The specific methane yield obtained in these runs was 357 dm3 kgVSfed−1 irrespective of OLR, and the methane content of biogas reached nearly 70%. These values are comparable with the methane yields from the mixture of sludge and hydrolysed SBP reported in the previous study (Borowski and Kucner, 2015). The process was unaffected by both ammonia and volatile fatty acids as discussed following. The literature dealing with the anaerobic digestion of stillage and sewage sludge is limited. Andalib et al. (2014) investigated an anaerobic fluidised bed bioreactor for digestion of primary municipal sludge and corn bioethanol thin stillage. The maximum methane production yields achieved in this study were 310 dm3 kgCOD−1 and 250 dm3 kgCOD−1, for sludge and stillage, respectively, which corresponded to around 620 dm3 kgVSfed−1 and 240 dm3 VSfed−1, respectively.
Behaviour of ammonia and volatile fatty acids
The characteristics of digestate from the experiments is depicted in Table 4, whereas the variations of TVFA, ammonium nitrogen and pH measured in the course of the runs are illustrated in Figures 3 and 4.
Characteristics of digestate.
SBPS: sugar beet pulp stillage; PM: poultry manure; MSS: municipal sewage sludge; TS: total solids; VS: volatile solids; TAN: total ammonium nitrogen; TVFA: total volatile fatty acids; TAL: total alkalinity.

Changes in pH, volatile fatty acids and ammonium nitrogen concentrations during the anaerobic co-digestion of stillage with municipal sewage sludge (R2-stages 1–2).
In the first and second stages of R1 the digester showed high instability evidenced by a rapid volatile fatty acids increase. The TVFA concentration reached nearly 12,000 g m−3 in R1-stage 1, whereas the volatile fatty acid to alkalinity ratio (TVFA/TAL), which indicates digester instability, increased to 1.15, which was much greater that a threshold inhibitory value of 0.4 (Kafle and Kim, 2013; Montanes et al., 2015). In stage 2 of R1, despite a 10% addition of poultry manure, these indicators even increased to nearly 17,000 g m−3 and 1.25, respectively (Figure 3). As mentioned above, a 20% addition of poultry manure to the SBP stillage greatly improved anaerobic digestion, and methane yields obtained in stages 3 and 4 of R1 were greater than the values reported in the literature. Interestingly, the digester was operated at high volatile fatty acids levels. The average TVFA concentration reported in R1-stage 3 was nearly 5500 g m−3, whereas the TVFA/TAL ratio averaged 0.39, close to the threshold of inhibition (Figure 4). In stage 4 the TVFA concentrations exceeded 8600 g m−3, whereas the average TVFA/TAL ratio was 0.59. The results confirm the previous findings according to which stable digestion operations are possible at TVFA concentrations of up to 10,000 g m−3 (Appels et al., 2008).
Considering the volatile fatty acid profile (Figure 5), it showed the dominance of propionic acid regarded as the main inhibitor of anaerobic digestion. The average propionic acid concentration measured in R1-stage 3 was 4145 g m−3, whereas the reported inhibitory concentrations of this acid are between 900 and 1500 g m−3 (Eskicioglu et al., 2011; Ma et al., 2009; Wang et al., 2009). Also, the propionic to acetic acid ratio of 3.7 was much higher than a value of 1.4 at which an inhibition of anaerobic digestion may occur (Nielsen et al., 2007). In stage 4 of R1 operated at the highest OLR of 5.24 kgVS m−3 d−1, the concentration of propionic acid increased to 6150 g m−3, however the drop in methane yield compared with stage 3 was insignificant (p = 0.3). Hence, it might be concluded that the high propionic acid level reported in co-digestion of stillage with poultry manure resulted in only a slight inhibition of methanogenesis, which is contradictory with the findings of other authors reporting the fate of this acid in anaerobic digestion. Furthermore, propionic acid has also been recognised as a product of ammonia inhibition (and vice versa – Banks et al., 2012; Moestedt et al., 2013), however the levels of free ammonia reported in our study were generally low with the average values of 112 and 84 g m−3 for stages 3 and 4 of R1, respectively. Although the inhibition of methanogenesis may appear at free ammonia concentration of 100 g m−3 (Eskicioglu et al., 2011) other authors reported stable digester operations at much higher values of 1000 g m−3 and more (Hansen et al., 1998). The low free ammonia levels can be attributed to the high VFA contents as discussed above, which reduced pH of the digestate.

Average concentrations of individual organic acids reported during the experiments.
Considering the co-digestion of stillage with sewage sludge performed in run R2, the concentrations of both ammonia and volatile fatty acids were far below the inhibitory levels for anaerobic digestion. Surprisingly, the average TVFA concentration reported in R2-stage 1 operated at lower OLR of 2.14 kgVS m−3 d−1 was nearly twice higher than the corresponding value detected in stage 2 of this run. The profiles of VFA in experiments R2 were markedly different from the ones for R1 discussed above. The main acid found in the digestate was acetic acid with the average concentrations of 1288 and 607 g m−3 in R2 trials, respectively. The average concentrations of propionic acid were 246 and 181 g m−3 for R2-stages 1 and 2, respectively, whereas the propionic to acetic acid ratio was far below 1.4. High concentrations of acetic acid within TVFA may indicate that acetoclastic methanogenesis is the rate-limiting step of anaerobic digestion treating the mixture of stillage and sewage sludge (Pagés-Díaz et al., 2015). As stillage is the product of SBP fermentation for bioethanol production, the role of lactic acid should also be discussed. The concentrations of lactic acid were generally lower than the values reported in our previous studies (Borowski and Kucner, 2015; Borowski et al., 2016), which could be attributed to the abundance of pentoses and raffinose with lower amounts of hexoses in stillage (Berłowska et al., 2017). As discussed previously, lactic acid is mainly produced from hexoses in fermentation processes, whereas other sugars that are not preferable substrates form most lactic acid bacteria (Abdel-Rahman et al., 2011; Castillo Martinez et al., 2013).
Conclusions
Semi-continuous digestion of SBPS alone is, in practise, infeasible because during the anaerobic treatment of this material a rapid acidification in the digester occurs leading to the inhibition of methane production. A solution to this problem may be the co-digestion with other organic wastes, including poultry manure or MSS. The highest methane production of nearly 420 dm3 kgVSfed−1 can be achieved from stillage mixed with 20% poultry manure, and the co-digestion process can successfully be performed even at high concentrations of propionic acid exceeding 4000 g m−3, which is contradictory with the literature findings. The co-digestion of stillage supplemented with 60% of sewage sludge can produce approximately 360 dm3 kgVSfed−1 with no inhibitory effects of both ammonia and volatile fatty acid. Hence, SBPS can be applied to enhance biogas production in existing anaerobic digestion installations located at municipal wastewater treatment plants.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research was financially supported by the National Centre for Research and Development (Poland) – project no. PBS1/B8/3/2012.
