- Email: [email protected]
Systematic comparison of mechanical and thermal sludge disintegration technologies
Waste Management 30 (2010) 1057–1062
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
Systematic comparison of mechanical and thermal sludge disintegration technologies B. Wett a,*, P. Phothilangka a, A. Eladawy b a b
Institute of Infrastructure, University of Innsbruck, Technikerstr. 13, 6020 Innsbruck, Austria Department of Civil Engineering, Helwan University, Elmataria, Cairo, Egypt
a r t i c l e
i n f o
Article history: Accepted 3 December 2009 Available online 8 January 2010
a b s t r a c t This study presents a systematic comparison and evaluation of sewage sludge pre-treatment by mechanical and thermal techniques. Waste activated sludge (WAS) was pre-treated by separate full scale Thermo-Pressure-Hydrolysis (TDH) and ball milling facilities. Then the sludge was processed in pilotscale digestion experiments. The results indicated that a significant increase in soluble organic matter could be achieved. TDH and ball milling pre-treatment could offer a feasible treatment method to efficiently disintegrate sludge and enhance biogas yield of digestion. The TDH increased biogas production by ca. 75% whereas ball milling allowed for an approximately 41% increase. The mechanisms of pre-treatment were investigated by numerical modeling based on Anaerobic Digestion Model No. 1 (ADM1) in the MatLab/SIMBA environment. TDH process induced advanced COD-solubilisation (CODsoluble/ CODtotal = 43%) and specifically complete destruction of cell mass which is hardly degradable in conventional digestion. While the ball mill technique achieved a lower solubilisation rate (CODsoluble/ CODtotal = 28%) and only a partial destruction of microbial decay products. From a whole-plant prospective relevant release of ammonia and formation of soluble inerts have been observed especially from thermal hydrolysis. Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction Sludge disintegration means pre-treatment technologies developed for both, accelerated hydrolysis rates and advanced bioavailability of solids for the subsequent digestion process. Corresponding benefits of pre-treatment are reduced requirements in residence time and volume, respectively, and an enhanced biogas production. The overall improvement of sludge stabilisation performance is supposed to reduce the final volatile solids concentration and viscosity for increased dewaterability and substantial cost savings for biosolids disposal. In general, three different process principles can be distinguished – chemical, biological and physical disintegration. The first two options require addition of chemicals like ozone (e.g. Weemaes et al., 2000), alkalinity or acid or enzymatic constituents (e.g. Barjenbruch and Kopplow, 2003). Within the third field of technologies basically heat, pressure or shear force are applied. These impacts can be achieved by sonication (e.g. Tiehm et al., 1997), thermal energy input or mechanical treatment (e.g. Hwang and Shin, 1997; Nah et al., 2000) for both rupturing effects and cavitations. Among these sludge disintegration technologies, two pre-treatment processes have been selected for a comparative study. * Corresponding author. Tel.: +43 660 811 4722; fax: +43 512 507 2911. E-mail address: [email protected] (B. Wett). 0956-053X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.wasman.2009.12.011
Thermal pre-treatment is a disintegration technique which has been installed in many wastewater treatment plants (WWTPs) (Kepp et al., 2000; Panter, 1998). The principle of this technique is that sludge cells are destructed by applying high-temperature optionally together with high pressure. Higher degradation efficiency is associated with higher biogas production and a lower content of volatile solids in the digested sludge which represents a smaller output of stabilized sludge with better dewatering properties (Zabranska et al., 2006). Kepp et al. (2000) implemented a full scale thermal pre-hydrolysis at the WWTP in Hamar, Norway, to pre-treat mixed raw-sludge with a temperature range of 165– 180 °C. Although digester volume for pre-treated sludge was smaller than for conventional digestion (1500 m3 and 3500 m3, respectively), the energy yield from biogas production increased from 678 kW (conventional digestion) to 1000 kW by thermal pre-treatment when the digester feed was 6–9 tons dissolved solids per day. Thermal pre-treatment is supposed to reduce also the hardly degradable materials (Elliott and Mahmood, 2007), therefore it improves the overall removal efficiency of organics in the digestion process. Ball milling treatment using small diameter balls is identified as an efficient solids rupturing technique for breaking up microbial cells to release intercellular materials (Elliott and Mahmood, 2007). Baier and Schmidheiny (1997) chose wet milling as the most promising mechanical pre-treatment option for WAS. Ball
1058
B. Wett et al. / Waste Management 30 (2010) 1057–1062
milling consistently showed better disintegration results than high speed cutter milling. However, solubilisation of organic material in WAS does not necessarily go in parallel with enhanced biogas yield. Ball milling pre-treated sludge showed only an increase of 10% in biogas production although volatile solid reduction was increased from 40% to 60%. The authors attributed this to the low rate digestion of cell disruption products which does not occur within 500 h. The specific break-down mechanisms of ball milling pre-treatment need closer investigation. In this paper, TDH and ball milling have been selected and applied to the same source of sludge for a direct and systematic comparison. Based on this case study, specific break-down mechanisms targeting different COD-fractions are investigated. Modelling degradation performance of acknowledged individual COD-fractions serves as a generic approach for technology evaluation. 2. Methods Fig. 1. Full scale ball milling installation.
2.1. Feed sludge To compare the impacts of mechanical and thermal pre-treatment, the same source of sludge was used for the experiments. Low loaded waste activated sludge (LLWAS) was collected from Strass WWTP treating almost exclusively municipal wastewater of maximum 200000 PE. The WWTP is operated with a two stage treatment called A and B process. Both stages represent completely separated sludge systems and produces high loaded waste activated sludge (HLWAS) and low loaded WAS (LLWAS), respectively (Böhnke, 1994). About 55–65% of the organic load is removed by the A-stage mainly based on adsorption and not on degradation (only ca. 12 h of sludge retention time SRT). The A-stage sludge shows similar degradability as primary sludge and therefore was not object to pre-treatment. Nitrogen removal in the downstream B-stage pre-denitrification system amounts to ca. 80% at a SRT of approximately 10 days. The characteristics of WAS are shown in Table 1.
Fig. 2. Full scale Thermo-Pressure-Hydrolysis (TDH) set-up.
2.2. Full scale disintegration
2.3. Pilot-scale digestion
Large scale installations for ball milling and thermal hydrolysis have been applied for pre-treatment of LLWAS from Strass WWTP.
Four anaerobic continuously stirred tank reactors (CSTR) with a liquid volume of 75 L each were used for digestion tests (Fig. 3, left). Two types of pre-treatment digestion experiments have been conducted in parallel in the reactors under mesophilic conditions (37 °C). After inoculation by digested WAS and an idle period of a few days, ball milling WAS was fed to reactor 1 and conventional WAS to reactor 2 at a feeding rate of 3.75 L/d for 35 days (at organic loading rate of 2.32 g COD/L/d and hydraulic residence time HRT of 20 days). Reactors 3 and 4 were loaded with TDH-WAS and conventional WAS, respectively, with the feeding rate of 5.4 L/day for 22 days (at organic loading rate of 3.34 g COD/L/d and HRT of 14 days). In fact, a prior study (Phothilangka et al., 2007) on the thermal pre-hydrolysis of WAS indicated that digestion with 14 and 20 days HRT yielded a similar amount of biogas. The operating pH-level in the reactors was in the range of 7.2–7.5. To investigate the transformation processes, nitrogen- (Hach-Lange-Photometry), carbon- (IR-sensitive C-S-analyser) and COD-compounds (HachLange-Photometry) were monitored and biogas quality (gas chromatography) and quantity (water displacement counter) were measured.
2.2.1. Ball milling The disintegration process was continuously performed with agitator ball mill Model LME 50 K (Fig. 1) applying fine sand balls (0.5–0.9 mm of diameter and 2.7 kg/L of density). The disintegration unit of ball milling consumes energy of 55 kW with a specific demand of 0.49 kWh/kg TSS. 2.2.2. Thermo-Pressure-Hydrolysis (TDH) TDH represents a continuously operated system involving high pressure pump, controlled pressure release valve and heat exchangers (Fig. 2). The WAS loading rate was set to 1.5 m3/h. The disintegration process was operated with a pressure of 19– 21 bar and feeding sludge was continuously pre-treated at 170– 180 °C for 60 min. Operating temperature below this range achieved less solubilisation (solubilisation rate only ca. 30% at 160 °C) and higher temperature promotes soluble inerts formation (Phothilangka, 2008).
Table 1 Average characteristics of low loaded activated sludge (LLWAS) used in experiments.
Value
TS (g/L)
VS (g/L)
CODt (mg/L)
CODs (mg/L)
TC (g/L)
TKN (mg/L)
NH4–N (mg/L)
pH
52.5
35.3
46378
3425
18.6
2296
264
7.6
1059
B. Wett et al. / Waste Management 30 (2010) 1057–1062
Fig. 3. (Left) Experimental set-up of four digesters with a total volume of 100 L each for two types of pre-treatment digestion experiments running in parallel with conventional WAS digestion; (right) ADM1 digester model configuration edited in SIMBA simulator environment.
2.4. Numerical modeling
of solids and leading to solubilisation of organic matter. The digestion of pre-hydrolysed substrate induces an almost instantaneous gas production jumping from almost zero to about 100 L/d within 2 days after an idle period without feeding. The gas production in the reactor with untreated sludge increased from about 8 L/d to less than 60 L/d and it took about 1 week to establish stable conversion rates. Gas production profiles (Fig. 4, right) – both measured and simulated ones – fluctuate due to variations in substrate concentration and composition which has been sampled and measured twice a week (Fig. 4, left). The calculation of the specific gas yield based on average VSS load and steady state gas production results an enhancement from 247 L to 443 L per kg VSS load due to TDH pre-treatment (ca. 75% increase illustrated in Fig. 4, right). Then, the corresponding benefit of TDH is an improvement of cake’s solids content from 25.2% to 32.7% TSS (after on-site dewatering using a spiral press at constant polymer dosage). Both effects enhanced degradation of organic matter and improved cake’s solids content – promise a total reduction in sludge disposal costs of about 25%. In the numerical model more rapid and more complete degradation of TDH-treated sludge is represented by calibrated disintegration rate and disintegration factors. Biokinetic parameters of acetogenesis and methanogenesis showed no sensitivity. These findings are in line with common knowledge about hydrolysis being the bottle-neck of anaerobic sludge digestion. TDH process impacts mainly biomass and decay products (matching measured ammonia release) whereas particulate inerts Xi already contained in the raw wastewater are hardly converted (Fig. 5). However, a side-effect of advanced protein degradation means ammonia release in digested sludge liquors. TDH induced a 64% NH4–N increase (from 1431 mg/L to 2241 mg/L) which causes a substantial internal load to the main treatment train. Measured
IWA’s Anaerobic Digestion Model ADM1 (Batstone et al., 2002) was applied. A configuration of ADM1 model digesters processing each type of feeding sludge were setup by using the Matlab-Simulink based commercial simulator SIMBA (Fig. 3, right). A model modification that includes an Xp-fraction (inactivated aerobic organisms and decay products) similar to the conventions of the activated sludge model series (ASM) improves the description of pre-treatment mechanisms significantly. Since thermal hydrolysis combined with anaerobic digestion targets mainly cell mass the add-up of the Xp-fraction is paramount especially for appropriate prediction of ammonia release (Wett et al., 2006). Detailed information on influent and raw-sludge characteristics and data from comparative digestion studies for pre-treated and untreated sludge have been used to calibrate a numerical model based on a carbon and nitrogen mass balance concept according to the principle of mass conservation i.e. particulate substrate compounds defined in ADM1 (proteins, lipids and carbo-hydrates) have not been directly measured but indirectly estimated from their mean nitrogen and carbon content. Mass flow spread sheets supporting the calibration procedure and resulting model parameter sets are presented elsewhere (Phothilangka, 2008).
3. Results and discussion 3.1. Thermal disintegration/TDH pre-hydrolysis Fig. 4 (left) shows the measured ratio of soluble COD to total COD which represents the actual hydrolysis efficiency. The comparison of untreated WAS and TDH-treated sludge indicates the strong impact of thermal disintegration by disrupting the structure
WAS
mea_WAS
TDH-S
m ea_TDHS
sim _WAS
sim_TDHS
120
Biogas production [L/d]
CODs/CODt [%]
50 40 30 20 10 0
100 80 60 40 20 0
5
8
11
14
17 20 Time [d]
23
26
29
32
3
6
9
12
15
18
21
24
27
30
33
Time [d]
Fig. 4. Comparison of conventional WAS and TDH-sludge in terms of hydrolysis efficiency (soluble COD to total COD ratio, left) and biogas production (right, continuous lines and dots shows model result and experimental results, respectively).
1060
B. Wett et al. / Waste Management 30 (2010) 1057–1062
Fig. 5. COD-fractionation of composite solids Xc of WAS with and without TDH pre-treatment (carbo-hydrates Xch, proteins Xpr, lipids Xli, decay products Xp, particulate inerts Xi, soluble inerts Si).
high ammonia increase due to TDH-pre-treatment supports the hypothesis of almost complete degradation of decay products. Inactivated aerobic biomass and decay products show a content in nitrogen of about 6% from proteins contained in cell mass. On the other hand, the particulate inerts Xi as part of raw sewage show a substantially lower share of nitrogen of about 2%. Nitrogen release figures indicate that this low nitrogen fraction is hardly degraded. The high soluble COD in the digester effluent indicates the generation of an additional by-product from high-temperature treatment. Model calibration revealed an increase of soluble inerts Si from 2% to 9% of Xc. Inert means non-degradable in applied treatment and no further degradability test have been conducted. Generated soluble inert compounds are diluted in the mainstream wastewater lane and contribute about 10 mg/L to the effluent COD.
ciency of ball milling treatment was mostly below 30% CODs/CODt ratio (Fig. 6, left) – a relatively low value compared to TDH-treatment achieving more than 40% hydrolysis. Biogas production of pre-treated sludge shows a sharp increase up to about 40 L/d while it remained around 30 for untreated sludge (overall increase by approximately 41%; Fig. 6, right). The steady state specific gas yield calculated from weekly measured VSS-loads was increased from 265 L to 415 L per kg VSS load for disintegrated WAS (DWAS). Although ball milling disrupted sludge cells and produced more soluble COD, the biogas yield was not increased as well as the increase of COD destruction rate. These results were in accordance with the studies of WAS pre-treatment by Baier and Schmidheiny (1997)which revealed that only some part of soluble COD was degraded and converted to biogas due to ball milling pre-treatment. Results from ADM1-calibration show the comparative fractionation of composite Xc of WAS in case with and without ball milling pre-treatment in Fig. 7. According to modeling results bioavailable substrates increased from 39% to 49% (Fig. 7) whereas the kinetic disintegration rate kdis of ball milling pre-treatment was switched from 0.25 to 1.0. The
3.2. Mechanical disintegration/ball mill treatment Secondary sludge from the same treatment plant was object to ball milling – a mechanical pre-treatment process. Hydrolysis effi-
WAS
sim_WAS
DWAS
35 Biogas production [L/d]
30 CODs/CODt [%]
measured_WAS
sim_DWAS
measured_DWAS
50
25 20 15 10 5 0
40 30 20 10 0
5
8
11
14
17 20 Time [d]
23
26
29
32
4
9
14
19
24 29 Time [d]
34
39
44
Fig. 6. Impacts of ball milling pre-treatment on the hydrolysis efficiency (left) and the increase of the biogas yield (right).
Fig. 7. COD-fractionation of composite solids Xc of WAS with and without ball milling pre-treatment.
49
1061
B. Wett et al. / Waste Management 30 (2010) 1057–1062 sim_WAS mea1_DWAS
CODt [mg/L]
50 45
mea1_WAS mea2_DWAS
Table 2 Summary of the main impacts of TDH compared with ball milling pre-treatment.
mea2_WAS sim_DWAS
average influent COD
40 35
39%degradation degradation 33%
30 25 20 15 10 5 0 0
5
10
15
20 25 Time[d]
30
35
40
45
Fig. 8. Simulated and measured CODtotal-profiles representing impacts on digestion performance.
dynamic increase of gas production from untreated sludge had been underestimated by the model but the dynamic drop in gas yield is predicted well. Both the corresponding biogas generation rate and the disintegration rate (compared to kdis = 1.5 of TDH pre-treatment) proved to be lower than for thermal pre-treatment. Therefore incomplete degradation during the HRT of 20 days resulted in a leftover of organic substances. Compared to TDH where Xp was completely degraded and disintegration rate was more efficiently accelerated, ball milling pre-treatment improved Xp-degradation by 26% (degraded from 38% to 28%). The simulated total COD matched well the measured COD as depicted in Fig. 8. The return load in ammonia and soluble inerts SI in sludge liquors showed less relevance, compared to TDH pre-treatment
sim_WAS
sim_DWAS
mea_WAS
mea_DWAS
1800
1600 NH4N [mg/L]
Biogas enhancement Xp-degradation Released NH4N Si generation
49% degradation 44% degradation
1400
1200
1000
800 0
25
50
75
100
125
150
175
200
Time[d] Fig. 9. Comparison of ammonia release profiles starting from high inoculum level.
TDH-WAS
Ball milling
+75% 100% +64% +387%
+41% 26% +29% +36%
(Fig. 9). The measured NH4–N concentration increased significantly due to TDH pre-treatment. Obviously the duration of the digestion experiment was too short to out-balance liquor concentrations. But the numerical model simulated an additional ammonia release of 29% increasing from a steady state value of 1084–1397 mg/L. The comparison of soluble inerts SI generated by ball milling pre-treatment and TDH process is illustrated in Fig. 10 and Table 2. The conversion factor for SI generated from composite solids XC has been calibrated to 2.0–2.5% for untreated WAS (represented by the baselines in Fig. 10). Insignificant SI increase due to ball milling could be described by a conversion factor fSI-XC of 3% whereas for TDH-treatment it had to be set to 9%. It seems that thermal treatment tends to produce inert organic soluble compounds and most of these by-products will be found in the plant effluent. Soluble compounds produced from mechanical disruption have been seen to be less resistant to biological degradation. 4. Conclusions Thermal hydrolysis proved to be more efficient than ball milling treatment in terms of positive and negative impacts on wholeplant performance: TDH increased biogas production by ca. 75% compared to approx. 41% by ball milling. Anaerobic degradation of organics in terms of total COD was enhanced from 33% to 44% in case of ball milling and to 51% due to TDH treatment. The generation of ammonia and inert soluble COD was substantially higher with thermal pre-treatment. The applied mechanistic model could identify the source of both products – additional biogas and released ammonia – in a more complete degradation of cell decay products. This finding explains why disintegration technologies are more effective on secondary sludge compared to primary sludge. Primary sludge shows already high degradability which can hardly be enhanced by pre-treatment techniques due to a lack of cell mass (less active biomass and almost no decay products compared to secondary sludge). Both investigated demonstration systems for sludge pre-treatment have not been optimized for energy reduction and recycling. Therefore results of this study cannot provide specific information on impacts on the overall energy balance – a crucial evaluation criterion for such techniques. In general thermal disintegration sys-
Ball milling WAS
WAS
WAS
5 Soluble inerts Si [kgCOD/m3]
5 Soluble inerts Si [kgCOD/m3]
TDH
4 3 2 1 0 0
20
40
60 Time[d]
80
100
4 3 2 1 0 0
20
40
60
80
100
Time [d]
Fig. 10. Simulated profiles of soluble inerts Si concentrations of digested WAS with TDH pre-treatment (left) and with ball milling pre-treatment (right).
1062
B. Wett et al. / Waste Management 30 (2010) 1057–1062
tems show more potential for recovery or efficient usage of waste energy while mechanical pre-treatment depends on electrical energy supply. Best-practice case studies on increasing numbers of implementations will support decision makers in appropriate technology selection.
References Baier, U., Schmidheiny, P., 1997. Enhanced anaerobic degradation of mechanically disintegrated sludge. Water Science and Technology 36 (11), 137–143. Barjenbruch, M., Kopplow, O., 2003. Encymatic, mechanical and thermal pretreatment of surplus sludge. Advances in Environmental Research 7 (3), 715–720. Batstone, D.J., Keller, J., Angelidaki, I., Kalyuzhnyi, S.V., Pavlostathis, S.G., Rozzi, A., Sanders, W.T.M., Siegrist, H., Vavilin, V.A., 2002. The IWA Anaerobic Digestion Model No. 1 (ADM1). Water Science and Technology 45 (10), 65–73. Böhnke, B., 1994. Nitrogen elimination in AB-treatment plants (in German). Korrespondenz Abwasser, 41, 6/94, S.900–907. Elliott, A., Mahmood, T., 2007. Pretreatment technologies for advancing anaerobic digestion of pulp and paper biotreatment residues. Water Research 41 (9), 4273–4286. Hwang, Y.K., Shin, B.E., 1997. A mechanical pretreatment of waste activated sludge for improvement of anaerobic digestion. Water Science and Technology 36 (12), 111–116.
Kepp, U., Machenbach, I., Weisz, N., Solheim, O.E., 2000. Enhanced stabilization of sewage sludge through thermal hydrolysis – three years of experience with full scale plant. Water Science and Technology 42 (9), 89–96. Nah, W.I., Kang, W.Y., Hwang, K., Song, W., 2000. Mechanical pretreatment of waste activated sludge for anaerobic digestion process. Water Research 34 (8), 2362– 2368. Panter, K., 1998. Cambi thermal hydrolysis – getting the bugs out of digestion and dewatering. In: Proceedings of 3rd European Biosolids and Organic Residuals Conference, Wakefield, England. Phothilangka, P., Schoen, A.M., Wett, B., 2007. Benefits and drawbacks of thermal pre-hydrolysis for operational performance of wastewater treatment plants. In: Proceedings of 5th International Symposium on Anaerobic Digestion of Solid Wastes and Energy Crops, Hammamet, Tunisia. Phothilangka, P., 2008. Sludge disintegration technologies for improved biogas yield. Ph.D. Thesis, Institut of Infrastructure, University of Innsbruck, Austria. Tiehm, A., Nickel, K., Neis, U., 1997. The use of ultrasound to accelerate the anaerobic digestion of sewage sludge. Water Science and Technology 36 (11), 121–128. Weemaes, M., Grootaerd, H., Simoens, F., Verstraete, W., 2000. Anaerobic digestion of ozonised biosolids. Water Research 34 (8), 2330–2336. Wett, B., Eladawy, A., Ogurek, M., 2006. Description of nitrogen incorporation and release in ADM1. Water Science and Technology 54 (4), 67–76. Zabranska, J., Dohanyos, M., Jenicek, P., Kutil, J., Cejka, J., 2006. Mechanical and rapid thermal disintegration methods of enhancement of biogas production-full-scale applications. In: Proceedings of IWA Specialized Conference on Sustainable Sludge Management: State of the Art, Challenges and Perspective, Moscow, Russia, pp. 235–241.
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
Systematic comparison of mechanical and thermal sludge disintegration technologies B. Wett a,*, P. Phothilangka a, A. Eladawy b a b
Institute of Infrastructure, University of Innsbruck, Technikerstr. 13, 6020 Innsbruck, Austria Department of Civil Engineering, Helwan University, Elmataria, Cairo, Egypt
a r t i c l e
i n f o
Article history: Accepted 3 December 2009 Available online 8 January 2010
a b s t r a c t This study presents a systematic comparison and evaluation of sewage sludge pre-treatment by mechanical and thermal techniques. Waste activated sludge (WAS) was pre-treated by separate full scale Thermo-Pressure-Hydrolysis (TDH) and ball milling facilities. Then the sludge was processed in pilotscale digestion experiments. The results indicated that a significant increase in soluble organic matter could be achieved. TDH and ball milling pre-treatment could offer a feasible treatment method to efficiently disintegrate sludge and enhance biogas yield of digestion. The TDH increased biogas production by ca. 75% whereas ball milling allowed for an approximately 41% increase. The mechanisms of pre-treatment were investigated by numerical modeling based on Anaerobic Digestion Model No. 1 (ADM1) in the MatLab/SIMBA environment. TDH process induced advanced COD-solubilisation (CODsoluble/ CODtotal = 43%) and specifically complete destruction of cell mass which is hardly degradable in conventional digestion. While the ball mill technique achieved a lower solubilisation rate (CODsoluble/ CODtotal = 28%) and only a partial destruction of microbial decay products. From a whole-plant prospective relevant release of ammonia and formation of soluble inerts have been observed especially from thermal hydrolysis. Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction Sludge disintegration means pre-treatment technologies developed for both, accelerated hydrolysis rates and advanced bioavailability of solids for the subsequent digestion process. Corresponding benefits of pre-treatment are reduced requirements in residence time and volume, respectively, and an enhanced biogas production. The overall improvement of sludge stabilisation performance is supposed to reduce the final volatile solids concentration and viscosity for increased dewaterability and substantial cost savings for biosolids disposal. In general, three different process principles can be distinguished – chemical, biological and physical disintegration. The first two options require addition of chemicals like ozone (e.g. Weemaes et al., 2000), alkalinity or acid or enzymatic constituents (e.g. Barjenbruch and Kopplow, 2003). Within the third field of technologies basically heat, pressure or shear force are applied. These impacts can be achieved by sonication (e.g. Tiehm et al., 1997), thermal energy input or mechanical treatment (e.g. Hwang and Shin, 1997; Nah et al., 2000) for both rupturing effects and cavitations. Among these sludge disintegration technologies, two pre-treatment processes have been selected for a comparative study. * Corresponding author. Tel.: +43 660 811 4722; fax: +43 512 507 2911. E-mail address: [email protected] (B. Wett). 0956-053X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.wasman.2009.12.011
Thermal pre-treatment is a disintegration technique which has been installed in many wastewater treatment plants (WWTPs) (Kepp et al., 2000; Panter, 1998). The principle of this technique is that sludge cells are destructed by applying high-temperature optionally together with high pressure. Higher degradation efficiency is associated with higher biogas production and a lower content of volatile solids in the digested sludge which represents a smaller output of stabilized sludge with better dewatering properties (Zabranska et al., 2006). Kepp et al. (2000) implemented a full scale thermal pre-hydrolysis at the WWTP in Hamar, Norway, to pre-treat mixed raw-sludge with a temperature range of 165– 180 °C. Although digester volume for pre-treated sludge was smaller than for conventional digestion (1500 m3 and 3500 m3, respectively), the energy yield from biogas production increased from 678 kW (conventional digestion) to 1000 kW by thermal pre-treatment when the digester feed was 6–9 tons dissolved solids per day. Thermal pre-treatment is supposed to reduce also the hardly degradable materials (Elliott and Mahmood, 2007), therefore it improves the overall removal efficiency of organics in the digestion process. Ball milling treatment using small diameter balls is identified as an efficient solids rupturing technique for breaking up microbial cells to release intercellular materials (Elliott and Mahmood, 2007). Baier and Schmidheiny (1997) chose wet milling as the most promising mechanical pre-treatment option for WAS. Ball
1058
B. Wett et al. / Waste Management 30 (2010) 1057–1062
milling consistently showed better disintegration results than high speed cutter milling. However, solubilisation of organic material in WAS does not necessarily go in parallel with enhanced biogas yield. Ball milling pre-treated sludge showed only an increase of 10% in biogas production although volatile solid reduction was increased from 40% to 60%. The authors attributed this to the low rate digestion of cell disruption products which does not occur within 500 h. The specific break-down mechanisms of ball milling pre-treatment need closer investigation. In this paper, TDH and ball milling have been selected and applied to the same source of sludge for a direct and systematic comparison. Based on this case study, specific break-down mechanisms targeting different COD-fractions are investigated. Modelling degradation performance of acknowledged individual COD-fractions serves as a generic approach for technology evaluation. 2. Methods Fig. 1. Full scale ball milling installation.
2.1. Feed sludge To compare the impacts of mechanical and thermal pre-treatment, the same source of sludge was used for the experiments. Low loaded waste activated sludge (LLWAS) was collected from Strass WWTP treating almost exclusively municipal wastewater of maximum 200000 PE. The WWTP is operated with a two stage treatment called A and B process. Both stages represent completely separated sludge systems and produces high loaded waste activated sludge (HLWAS) and low loaded WAS (LLWAS), respectively (Böhnke, 1994). About 55–65% of the organic load is removed by the A-stage mainly based on adsorption and not on degradation (only ca. 12 h of sludge retention time SRT). The A-stage sludge shows similar degradability as primary sludge and therefore was not object to pre-treatment. Nitrogen removal in the downstream B-stage pre-denitrification system amounts to ca. 80% at a SRT of approximately 10 days. The characteristics of WAS are shown in Table 1.
Fig. 2. Full scale Thermo-Pressure-Hydrolysis (TDH) set-up.
2.2. Full scale disintegration
2.3. Pilot-scale digestion
Large scale installations for ball milling and thermal hydrolysis have been applied for pre-treatment of LLWAS from Strass WWTP.
Four anaerobic continuously stirred tank reactors (CSTR) with a liquid volume of 75 L each were used for digestion tests (Fig. 3, left). Two types of pre-treatment digestion experiments have been conducted in parallel in the reactors under mesophilic conditions (37 °C). After inoculation by digested WAS and an idle period of a few days, ball milling WAS was fed to reactor 1 and conventional WAS to reactor 2 at a feeding rate of 3.75 L/d for 35 days (at organic loading rate of 2.32 g COD/L/d and hydraulic residence time HRT of 20 days). Reactors 3 and 4 were loaded with TDH-WAS and conventional WAS, respectively, with the feeding rate of 5.4 L/day for 22 days (at organic loading rate of 3.34 g COD/L/d and HRT of 14 days). In fact, a prior study (Phothilangka et al., 2007) on the thermal pre-hydrolysis of WAS indicated that digestion with 14 and 20 days HRT yielded a similar amount of biogas. The operating pH-level in the reactors was in the range of 7.2–7.5. To investigate the transformation processes, nitrogen- (Hach-Lange-Photometry), carbon- (IR-sensitive C-S-analyser) and COD-compounds (HachLange-Photometry) were monitored and biogas quality (gas chromatography) and quantity (water displacement counter) were measured.
2.2.1. Ball milling The disintegration process was continuously performed with agitator ball mill Model LME 50 K (Fig. 1) applying fine sand balls (0.5–0.9 mm of diameter and 2.7 kg/L of density). The disintegration unit of ball milling consumes energy of 55 kW with a specific demand of 0.49 kWh/kg TSS. 2.2.2. Thermo-Pressure-Hydrolysis (TDH) TDH represents a continuously operated system involving high pressure pump, controlled pressure release valve and heat exchangers (Fig. 2). The WAS loading rate was set to 1.5 m3/h. The disintegration process was operated with a pressure of 19– 21 bar and feeding sludge was continuously pre-treated at 170– 180 °C for 60 min. Operating temperature below this range achieved less solubilisation (solubilisation rate only ca. 30% at 160 °C) and higher temperature promotes soluble inerts formation (Phothilangka, 2008).
Table 1 Average characteristics of low loaded activated sludge (LLWAS) used in experiments.
Value
TS (g/L)
VS (g/L)
CODt (mg/L)
CODs (mg/L)
TC (g/L)
TKN (mg/L)
NH4–N (mg/L)
pH
52.5
35.3
46378
3425
18.6
2296
264
7.6
1059
B. Wett et al. / Waste Management 30 (2010) 1057–1062
Fig. 3. (Left) Experimental set-up of four digesters with a total volume of 100 L each for two types of pre-treatment digestion experiments running in parallel with conventional WAS digestion; (right) ADM1 digester model configuration edited in SIMBA simulator environment.
2.4. Numerical modeling
of solids and leading to solubilisation of organic matter. The digestion of pre-hydrolysed substrate induces an almost instantaneous gas production jumping from almost zero to about 100 L/d within 2 days after an idle period without feeding. The gas production in the reactor with untreated sludge increased from about 8 L/d to less than 60 L/d and it took about 1 week to establish stable conversion rates. Gas production profiles (Fig. 4, right) – both measured and simulated ones – fluctuate due to variations in substrate concentration and composition which has been sampled and measured twice a week (Fig. 4, left). The calculation of the specific gas yield based on average VSS load and steady state gas production results an enhancement from 247 L to 443 L per kg VSS load due to TDH pre-treatment (ca. 75% increase illustrated in Fig. 4, right). Then, the corresponding benefit of TDH is an improvement of cake’s solids content from 25.2% to 32.7% TSS (after on-site dewatering using a spiral press at constant polymer dosage). Both effects enhanced degradation of organic matter and improved cake’s solids content – promise a total reduction in sludge disposal costs of about 25%. In the numerical model more rapid and more complete degradation of TDH-treated sludge is represented by calibrated disintegration rate and disintegration factors. Biokinetic parameters of acetogenesis and methanogenesis showed no sensitivity. These findings are in line with common knowledge about hydrolysis being the bottle-neck of anaerobic sludge digestion. TDH process impacts mainly biomass and decay products (matching measured ammonia release) whereas particulate inerts Xi already contained in the raw wastewater are hardly converted (Fig. 5). However, a side-effect of advanced protein degradation means ammonia release in digested sludge liquors. TDH induced a 64% NH4–N increase (from 1431 mg/L to 2241 mg/L) which causes a substantial internal load to the main treatment train. Measured
IWA’s Anaerobic Digestion Model ADM1 (Batstone et al., 2002) was applied. A configuration of ADM1 model digesters processing each type of feeding sludge were setup by using the Matlab-Simulink based commercial simulator SIMBA (Fig. 3, right). A model modification that includes an Xp-fraction (inactivated aerobic organisms and decay products) similar to the conventions of the activated sludge model series (ASM) improves the description of pre-treatment mechanisms significantly. Since thermal hydrolysis combined with anaerobic digestion targets mainly cell mass the add-up of the Xp-fraction is paramount especially for appropriate prediction of ammonia release (Wett et al., 2006). Detailed information on influent and raw-sludge characteristics and data from comparative digestion studies for pre-treated and untreated sludge have been used to calibrate a numerical model based on a carbon and nitrogen mass balance concept according to the principle of mass conservation i.e. particulate substrate compounds defined in ADM1 (proteins, lipids and carbo-hydrates) have not been directly measured but indirectly estimated from their mean nitrogen and carbon content. Mass flow spread sheets supporting the calibration procedure and resulting model parameter sets are presented elsewhere (Phothilangka, 2008).
3. Results and discussion 3.1. Thermal disintegration/TDH pre-hydrolysis Fig. 4 (left) shows the measured ratio of soluble COD to total COD which represents the actual hydrolysis efficiency. The comparison of untreated WAS and TDH-treated sludge indicates the strong impact of thermal disintegration by disrupting the structure
WAS
mea_WAS
TDH-S
m ea_TDHS
sim _WAS
sim_TDHS
120
Biogas production [L/d]
CODs/CODt [%]
50 40 30 20 10 0
100 80 60 40 20 0
5
8
11
14
17 20 Time [d]
23
26
29
32
3
6
9
12
15
18
21
24
27
30
33
Time [d]
Fig. 4. Comparison of conventional WAS and TDH-sludge in terms of hydrolysis efficiency (soluble COD to total COD ratio, left) and biogas production (right, continuous lines and dots shows model result and experimental results, respectively).
1060
B. Wett et al. / Waste Management 30 (2010) 1057–1062
Fig. 5. COD-fractionation of composite solids Xc of WAS with and without TDH pre-treatment (carbo-hydrates Xch, proteins Xpr, lipids Xli, decay products Xp, particulate inerts Xi, soluble inerts Si).
high ammonia increase due to TDH-pre-treatment supports the hypothesis of almost complete degradation of decay products. Inactivated aerobic biomass and decay products show a content in nitrogen of about 6% from proteins contained in cell mass. On the other hand, the particulate inerts Xi as part of raw sewage show a substantially lower share of nitrogen of about 2%. Nitrogen release figures indicate that this low nitrogen fraction is hardly degraded. The high soluble COD in the digester effluent indicates the generation of an additional by-product from high-temperature treatment. Model calibration revealed an increase of soluble inerts Si from 2% to 9% of Xc. Inert means non-degradable in applied treatment and no further degradability test have been conducted. Generated soluble inert compounds are diluted in the mainstream wastewater lane and contribute about 10 mg/L to the effluent COD.
ciency of ball milling treatment was mostly below 30% CODs/CODt ratio (Fig. 6, left) – a relatively low value compared to TDH-treatment achieving more than 40% hydrolysis. Biogas production of pre-treated sludge shows a sharp increase up to about 40 L/d while it remained around 30 for untreated sludge (overall increase by approximately 41%; Fig. 6, right). The steady state specific gas yield calculated from weekly measured VSS-loads was increased from 265 L to 415 L per kg VSS load for disintegrated WAS (DWAS). Although ball milling disrupted sludge cells and produced more soluble COD, the biogas yield was not increased as well as the increase of COD destruction rate. These results were in accordance with the studies of WAS pre-treatment by Baier and Schmidheiny (1997)which revealed that only some part of soluble COD was degraded and converted to biogas due to ball milling pre-treatment. Results from ADM1-calibration show the comparative fractionation of composite Xc of WAS in case with and without ball milling pre-treatment in Fig. 7. According to modeling results bioavailable substrates increased from 39% to 49% (Fig. 7) whereas the kinetic disintegration rate kdis of ball milling pre-treatment was switched from 0.25 to 1.0. The
3.2. Mechanical disintegration/ball mill treatment Secondary sludge from the same treatment plant was object to ball milling – a mechanical pre-treatment process. Hydrolysis effi-
WAS
sim_WAS
DWAS
35 Biogas production [L/d]
30 CODs/CODt [%]
measured_WAS
sim_DWAS
measured_DWAS
50
25 20 15 10 5 0
40 30 20 10 0
5
8
11
14
17 20 Time [d]
23
26
29
32
4
9
14
19
24 29 Time [d]
34
39
44
Fig. 6. Impacts of ball milling pre-treatment on the hydrolysis efficiency (left) and the increase of the biogas yield (right).
Fig. 7. COD-fractionation of composite solids Xc of WAS with and without ball milling pre-treatment.
49
1061
B. Wett et al. / Waste Management 30 (2010) 1057–1062 sim_WAS mea1_DWAS
CODt [mg/L]
50 45
mea1_WAS mea2_DWAS
Table 2 Summary of the main impacts of TDH compared with ball milling pre-treatment.
mea2_WAS sim_DWAS
average influent COD
40 35
39%degradation degradation 33%
30 25 20 15 10 5 0 0
5
10
15
20 25 Time[d]
30
35
40
45
Fig. 8. Simulated and measured CODtotal-profiles representing impacts on digestion performance.
dynamic increase of gas production from untreated sludge had been underestimated by the model but the dynamic drop in gas yield is predicted well. Both the corresponding biogas generation rate and the disintegration rate (compared to kdis = 1.5 of TDH pre-treatment) proved to be lower than for thermal pre-treatment. Therefore incomplete degradation during the HRT of 20 days resulted in a leftover of organic substances. Compared to TDH where Xp was completely degraded and disintegration rate was more efficiently accelerated, ball milling pre-treatment improved Xp-degradation by 26% (degraded from 38% to 28%). The simulated total COD matched well the measured COD as depicted in Fig. 8. The return load in ammonia and soluble inerts SI in sludge liquors showed less relevance, compared to TDH pre-treatment
sim_WAS
sim_DWAS
mea_WAS
mea_DWAS
1800
1600 NH4N [mg/L]
Biogas enhancement Xp-degradation Released NH4N Si generation
49% degradation 44% degradation
1400
1200
1000
800 0
25
50
75
100
125
150
175
200
Time[d] Fig. 9. Comparison of ammonia release profiles starting from high inoculum level.
TDH-WAS
Ball milling
+75% 100% +64% +387%
+41% 26% +29% +36%
(Fig. 9). The measured NH4–N concentration increased significantly due to TDH pre-treatment. Obviously the duration of the digestion experiment was too short to out-balance liquor concentrations. But the numerical model simulated an additional ammonia release of 29% increasing from a steady state value of 1084–1397 mg/L. The comparison of soluble inerts SI generated by ball milling pre-treatment and TDH process is illustrated in Fig. 10 and Table 2. The conversion factor for SI generated from composite solids XC has been calibrated to 2.0–2.5% for untreated WAS (represented by the baselines in Fig. 10). Insignificant SI increase due to ball milling could be described by a conversion factor fSI-XC of 3% whereas for TDH-treatment it had to be set to 9%. It seems that thermal treatment tends to produce inert organic soluble compounds and most of these by-products will be found in the plant effluent. Soluble compounds produced from mechanical disruption have been seen to be less resistant to biological degradation. 4. Conclusions Thermal hydrolysis proved to be more efficient than ball milling treatment in terms of positive and negative impacts on wholeplant performance: TDH increased biogas production by ca. 75% compared to approx. 41% by ball milling. Anaerobic degradation of organics in terms of total COD was enhanced from 33% to 44% in case of ball milling and to 51% due to TDH treatment. The generation of ammonia and inert soluble COD was substantially higher with thermal pre-treatment. The applied mechanistic model could identify the source of both products – additional biogas and released ammonia – in a more complete degradation of cell decay products. This finding explains why disintegration technologies are more effective on secondary sludge compared to primary sludge. Primary sludge shows already high degradability which can hardly be enhanced by pre-treatment techniques due to a lack of cell mass (less active biomass and almost no decay products compared to secondary sludge). Both investigated demonstration systems for sludge pre-treatment have not been optimized for energy reduction and recycling. Therefore results of this study cannot provide specific information on impacts on the overall energy balance – a crucial evaluation criterion for such techniques. In general thermal disintegration sys-
Ball milling WAS
WAS
WAS
5 Soluble inerts Si [kgCOD/m3]
5 Soluble inerts Si [kgCOD/m3]
TDH
4 3 2 1 0 0
20
40
60 Time[d]
80
100
4 3 2 1 0 0
20
40
60
80
100
Time [d]
Fig. 10. Simulated profiles of soluble inerts Si concentrations of digested WAS with TDH pre-treatment (left) and with ball milling pre-treatment (right).
1062
B. Wett et al. / Waste Management 30 (2010) 1057–1062
tems show more potential for recovery or efficient usage of waste energy while mechanical pre-treatment depends on electrical energy supply. Best-practice case studies on increasing numbers of implementations will support decision makers in appropriate technology selection.
References Baier, U., Schmidheiny, P., 1997. Enhanced anaerobic degradation of mechanically disintegrated sludge. Water Science and Technology 36 (11), 137–143. Barjenbruch, M., Kopplow, O., 2003. Encymatic, mechanical and thermal pretreatment of surplus sludge. Advances in Environmental Research 7 (3), 715–720. Batstone, D.J., Keller, J., Angelidaki, I., Kalyuzhnyi, S.V., Pavlostathis, S.G., Rozzi, A., Sanders, W.T.M., Siegrist, H., Vavilin, V.A., 2002. The IWA Anaerobic Digestion Model No. 1 (ADM1). Water Science and Technology 45 (10), 65–73. Böhnke, B., 1994. Nitrogen elimination in AB-treatment plants (in German). Korrespondenz Abwasser, 41, 6/94, S.900–907. Elliott, A., Mahmood, T., 2007. Pretreatment technologies for advancing anaerobic digestion of pulp and paper biotreatment residues. Water Research 41 (9), 4273–4286. Hwang, Y.K., Shin, B.E., 1997. A mechanical pretreatment of waste activated sludge for improvement of anaerobic digestion. Water Science and Technology 36 (12), 111–116.
Kepp, U., Machenbach, I., Weisz, N., Solheim, O.E., 2000. Enhanced stabilization of sewage sludge through thermal hydrolysis – three years of experience with full scale plant. Water Science and Technology 42 (9), 89–96. Nah, W.I., Kang, W.Y., Hwang, K., Song, W., 2000. Mechanical pretreatment of waste activated sludge for anaerobic digestion process. Water Research 34 (8), 2362– 2368. Panter, K., 1998. Cambi thermal hydrolysis – getting the bugs out of digestion and dewatering. In: Proceedings of 3rd European Biosolids and Organic Residuals Conference, Wakefield, England. Phothilangka, P., Schoen, A.M., Wett, B., 2007. Benefits and drawbacks of thermal pre-hydrolysis for operational performance of wastewater treatment plants. In: Proceedings of 5th International Symposium on Anaerobic Digestion of Solid Wastes and Energy Crops, Hammamet, Tunisia. Phothilangka, P., 2008. Sludge disintegration technologies for improved biogas yield. Ph.D. Thesis, Institut of Infrastructure, University of Innsbruck, Austria. Tiehm, A., Nickel, K., Neis, U., 1997. The use of ultrasound to accelerate the anaerobic digestion of sewage sludge. Water Science and Technology 36 (11), 121–128. Weemaes, M., Grootaerd, H., Simoens, F., Verstraete, W., 2000. Anaerobic digestion of ozonised biosolids. Water Research 34 (8), 2330–2336. Wett, B., Eladawy, A., Ogurek, M., 2006. Description of nitrogen incorporation and release in ADM1. Water Science and Technology 54 (4), 67–76. Zabranska, J., Dohanyos, M., Jenicek, P., Kutil, J., Cejka, J., 2006. Mechanical and rapid thermal disintegration methods of enhancement of biogas production-full-scale applications. In: Proceedings of IWA Specialized Conference on Sustainable Sludge Management: State of the Art, Challenges and Perspective, Moscow, Russia, pp. 235–241.