This is the accepted version of the journal article: Guerra-Gorostegi, N.; González, D.; Puyuelo, B.; [et al.]. «Biomass fuel produc- tion from cellulosic sludge through biodrying : Aeration strategies, quality of end-products, gaseous emissions and techno-economic assessment». Waste Man- agement, Vol. 126 (May 2021), p. 487-496. DOI 10.1016/j.wasman.2021.03.036 This version is available at https://ddd.uab.cat/record/264029 under the terms of the license 1 Biomass fuel production from cellulosic sludge through biodrying: aeration 2 strategies, quality of end-products, gaseous emissions and techno-economic 3 assessment 4 5 Guerra- Gorostegi, N.a, González, D.b,c ¸Puyuelo, B. a, Ovejero, J.a, Colón, J.a, Gabriel, 6 D.c, Sánchez, A.b, Ponsá, Sa*. a 7 BETA Technological Center, Science and Technology Faculty, University of Vic- 8 Central University of Catalonia. 08500 Vic, Barcelona, Spain b 9 Composting Research Group (GICOM), Dept. of Chemical, Biological and 10 Environmental Engineering, Universitat Autònoma de Barcelona, 08193 Bellaterra, 11 Barcelona, Spain c 12 Group of Biological Treatment of Liquid and Gaseous Effluents (GENOCOV), Dept. of 13 Chemical, Biological and Environmental Engineering, Universitat Autònoma de 14 Barcelona, 08193 Bellaterra, Barcelona, Spain 15 * Corresponding author: 16 Sergio Ponsá 17 BETA Technological Center, 18 Science and Technology Faculty, 19 University of Vic-Central University of Catalonia 20 08500 Vic, Barcelona, Spain 21 Email address: sergio.ponsa@uvic.cat 1 22 Abstract 23 This study assesses the technological, environmental and economic feasibility of 24 biodrying to valorise cellulosic sludge as renewable energy source. Specifically, three 25 different aeration strategies were compared in terms of biodrying performance, 26 energetic consumption, gaseous emissions, quality of end-products and techno- 27 economic analysis. The overall drying efficiency of the process was evaluated through 3 28 different aeration strategies. Those strategies were based on different combinations of 29 convective drying with biogenic heat produced. Two innovative biodrying performance 30 indicators (Energetic Biodrying Index and Biodrying Performance Index) are proposed 31 to better assess the initial and operational conditions that are favouring the maximum 32 energy process efficiency and the highest quality of the end-products. The end-products 33 obtained always presented moisture contents below 40% and lower heating values 34 above 9.4 MJ·kg-1, being the best values achieved, 32.6% and 10.4 MJ·kg-1 for moisture 35 content and lower heating value respectively. Low N2O and CH4 emissions confirmed 36 the effective aeration during all the three strategies carried out, while the rest of the 37 gases monitored were related either to temperature or biological phenomena. A techno- 38 economic analysis proved the economic viability and attractiveness of the biodrying 39 technology for cellulosic sludge in all the strategies applied. 40 Keywords: cellulosic sludge, biodrying, aeration strategies, gaseous emissions, techno- 41 economic analysis. 42 43 44 2 45 Abbreviation list: AT4 4 days cumulative oxygen consumption B.A. Bulking Agent B.I. Biodrying Index B.P.I. Biodrying Performance Index CAPEX Capital expenditures COD Chemical Oxygen Demand C.S. Cellulosic Sludge D.R.I. Dynamic Respirometric Index E.B.I. Energetic Biodrying Index E.C. Energy Consumption E.P. Energy Production GHG Greenhouse Gases FAS Free Air Space HHV Higher Heating Value IRR Internal Rate of Return LHV Lower Heating Value MC Moisture Content MSW Municipal Solid Waste NPV Net Present value OPEX Operational expenditures SRF Solid Recovered Fuels TIP Temperature Increasing Phase TS Total Solids tVOC Total Volatile Organic Compounds VS Volatile Solids VS-CS Volatile Solids from Cellulosic Sludge WWTP Wastewater Treatment Plant 46 47 3 48 1. Introduction 49 The recovery of resources (materials and energy) from wastewater is a promising 50 solution that contributes to achieve the relevant sustainability challenges of water 51 utilities and modern societies. There is a wide range of innovative technologies that are 52 being applied that additionally make the wastewater treatment plants more efficient; 53 reducing the amount of sludge produced, reducing the energy consumed and clearly 54 providing environmental and economic benefits (Conca et al., 2020; Da Ros et al., 55 2020). These new technologies are applied at different stages of the water treatment 56 facilities, but mostly in side streams and down streams. On the one hand, these flows 57 present high concentrations of COD, TS and nutrients and are normally considered 58 suitable candidates to implement resource recovery strategies (Raheem et al., 2018). 59 Nevertheless, the impact on the overall improvement of the Wastewater Treatment Plant 60 (WWTP) efficiency and energy savings, although it is positive, still has some 61 limitations and margin for improvement. 62 To increase the WWTP efficiency while increasing the resource recovery capacity, 63 new technologies are developed and applied at the first stages of the main stream 64 (Reijken et al., 2018; Larriba et al., 2020). A part of the direct impacts on resource 65 recovery, these technologies have also an indirect impact on improving the efficiency of 66 the next stages since they significantly reduce the chemical oxygen demand (COD) and 67 total solids (TS) content and consequently, reducing the aeration needs and the amounts 68 of sewage sludge produced that can be translated in important energy savings. 69 Among these innovative promising technologies, Cellvation® is aiming to maximise 70 the recovery and recycling of cellulose replacing, partially or totally, the primary settler. 4 71 Cellvation®, consists of an initial grit and hair removal in a rotating drum filter 72 followed by a 350 µm fine sieve (Salsnes Filter, Norway), followed by a cellpress, a 73 hygienisation step and a pelletiser. Finally, cellulose is recovered in the form of Recell® 74 cellulose pellets. However, a cellulosic sludge (CS) is also produced. To avoid the loss 75 of resources and minimise disposal costs, the CS could be further valorised considering 76 the high potential energetic content of this material due to the high content in cellulose 77 and hemicellulose. Thus, all the cellulose-rich sludge obtained after the cellpress, could 78 be directly considered as suitable material for being valorised via energy recovery 79 technologies. 80 Among the different technologies that could be applied, biodrying is presented as an 81 innovative, energy-saving and environmentally friendly alternative process for CS and 82 sewage sludge energetic valorisation. Biodrying, that can be considered as a 83 composting-like process, is an aerobic biological process that uses the biogenic heat 84 produced during the decomposition of biodegradable organic matter to remove as much 85 moisture as possible in the shortest operation time (Cai et al., 2012). Additionally, 86 biodrying is aiming to preserve, in the final biomass fuel produced, most of the organic 87 matter content present in the raw material (Huiliñir and Villegas, 2014). 88 Biodrying performance is normally assessed using two main indices: the daily drying 89 rate and Biodrying Index (BI). However, theses indices present some limitations, since 90 daily drying rates does not consider the organic carbon biodegraded and BI does not 91 consider the external energy consumed. 92 CS presents some difficulties over other organic wastes for its valorisation through 93 biodrying, being the most significant its low porosity and high moisture content, which 5 94 can hamper the proper air diffusion throughout the raw material. This technology has 95 not been optimized yet for low porosity organic wastes and there is still room for the 96 improvement of its performance and efficiency that would importantly increase its 97 possible applications to valorise different types of sewage sludge. 98 Assuming suitable initial conditions (e.g. organic content of raw materials and matrix 99 structure and porosity), water evaporation in the biodrying process depends mainly on 100 two operational parameters: (1) airflow temperature (inlet and outlet) and (2) airflow 101 rate. In literature, there are some studies assessing the performance of biodrying 102 processes using continuous and discontinuous aeration strategies and a wide range of 103 specific airflow rates from 0.5 to 6.2 L min-1kg-1VS-CS (Zhao et al., 2010; Huiliñir and 104 Villegas, 2014). 105 However, an effective biodrying process should not only be considered from a 106 technical perspective but also by its environmental and economic sustainability. 107 Therefore, an optimized biodrying process should guarantee: (1) low energy 108 consumption and low harmful gaseous emissions, allowing, in turn, (2) the production 109 of a high-quality biomass-fuel maximising the net energy recovery. The key quality 110 indicators of the biodried products obtained are a low moisture content (MC) and high 111 calorific potential. There are few previous references about economic viability of 112 biodrying technologies applied to low-porosity materials. In these studies, the main 113 weaknesses were indeed related to the high MC of final products (Navaee-Ardeh et al., 114 2006). In addition, electricity demand, particularly for aeration, is recognised to be the 115 main operational cost during biodrying processes (Psaltis and Komilis, 2019). 116 Consequently, choosing the most appropriate aeration strategy will lead to important 117 energy savings as well as the improved environmental performance of the process. 6 118 Regarding environmental performance, there also exists a lack of information in the 119 literature about gaseous emissions during biodrying processes, for both sewage sludge 120 and Municipal Solid Waste (MSW) valorisation (Ragazzi et al., 2011; González et al., 121 2019a). 122 Therefore, the main objective of this study was to develop an in-deep performance 123 assessment of biodrying from a technical, environmental and economic point of view in 124 the particular case of CS used as raw material. Specifically, process performances and 125 quality of end-products under three different aeration strategies were compared in terms 126 of process efficiency, gaseous emissions and economic feasibility. In addition, new 127 process performance indices are proposed to overcome the limitations of the currently 128 used indices and have a more detailed and comprehensive assessment of biodrying 129 processes. 130 2. Materials and methods 131 2.1. Raw materials and initial mixture 132 Cellulosic sludge was collected from the WWTP of Geestmerambacht, The 133 Netherlands. In this case, Cellvation® cellulose recovery technology treats 30-80 m3·h-1 134 of wastewater. This system reduces the total suspended solids up to 40%, that can be 135 translated into energy savings up to 15% and reduction of sewage sludge production up 136 to 20% (Cellvation, B.V., 2018). The raw material used in this study is a mix of 137 intermediate cellulose-rich flows, the so-called, CS. The main physic-chemical 138 characteristics of CS are presented in Table 1, including a comparison with other 139 conventional sludge produced. 7 140 Pruning waste was used as bulking agent (B.A). It was obtained from the MSW 141 composting plant of the Parc Ambiental de Bufalvent located in Manresa, Spain. 142 Sludge and bulking agent were mixed manually. The mixture ratio used was 1:2.5 of 143 CS to pruning waste, allowing to have a MC and Free Air Space (FAS) within the 144 optimal range, close to 50-60% (Villegas and Huiliñir, 2015) and close to 70%, 145 respectively, of all the initial mixtures used in this study. 146 2.2. Experimental equipment and operation 147 2.2.1 Biodrying reactor operation 148 A near-to-adiabatic reactor with a working volume of 100L was used for all 149 biodrying trials. The reactor was aerated through a diffusion grid in the bottom using an 150 air compressor (Dixair DNX 2050, Worthington Creyssensac) and a 151 flowmeter/controller (D-6311-DR, Bronkhorst High-Tech B.V.). Humidity of inlet air 152 was controlled by installing a set of 2 filters for moisture and particle removal before 153 the flow meter/controller. During the biodrying trials inlet air and matrix temperatures 154 were monitored using proves (Pt-100). A representative sample of exhaust gases (0.14 155 L·min-1) was continuously pumped and analysed using O2 and CO2 sensors (O2A2 and 156 IRC A1, respectively, Alphasense). Weight loss was monitored with a scale (Gram 157 Precision / k3-k3i, Gram group). Arduino UNO was used for data acquisition and 158 LabView2017 (National Instruments) software was used for data analysis, process 159 monitoring and airflow control. Material homogenization, was carried out using a maze 160 spiral compost aerator. The turning frequency criteria adopted was once per day during 161 the thermophilic stage of the process while it was once per two days during late 162 mesophilic and cooling stages. 8 163 2.2.2. Control system 164 Three different aeration strategies were adopted for cellulosic sludge biodrying: (1) 165 to reach and maintain the highest temperature of bulk material and the longest duration 166 of thermophilic phase (S1); (2) setting high airflows, tripling the values of S1 airflows 167 (S2); and (3) a combined strategy where S1 airflows were maintained until the 168 thermophilic phase was over (below 45oC), plus S2 airflows use thereafter (S3). An 169 algorithm to adapt aeration rates to 5 temperature ranges (<35ºC, 35-45ºC, 45-55ºC, 55- 170 70ºC and > 70ºC) was developed in which aeration rates per each range were adapted to 171 the particular strategy assessed. For the first strategy, optimal aeration levels typically 172 used during composting process were chosen, aiming bulk temperature to be the main 173 water removal force. For the second strategy, aeration levels were set to be significantly 174 higher, particularly in the thermophilic stage, in order to facilitate the extraction of the 175 evaporated water and ultimately to improve water removal (Navaee-Ardeh et al., 2006). 176 In the third trial, a combination of the previous strategies was tested aiming to maximise 177 moisture removal in the two stages by firstly maximising temperature (equivalent to S1) 178 during the thermophilic stage and secondly maximising aeration rates (equivalent to S2) 179 during mesophilic-cooling stage. 180 181 2.3. Analytical methods 182 All analysis were made following the Standard Methods for the Examination of 183 Water and Wastewater (APHA, 1995) except for pH and conductivity measurements, 184 which were carried out following the Test Methods for the Examination of Composting 185 and Compost (US Department of Agriculture and US Composting Council, 2001). C/N 9 186 and FAS values were estimated from chemical characterisation as suggested and used 187 elsewhere (Richard et al., 2002; Villegas & Huiliñir, 2014). Biological stability, by 188 means of Dynamic Respiration Index (DRI) and 4 days cumulative oxygen consumption 189 (AT4), were determined in a dynamic respirometer developed by Ponsa et al. (2010). 190 Higher calorific value (HHV) of wastes was determined using a bomb calorimeter 191 (1341 Plain Jacket Calorimeter with the 1108 Oxygen Combustion Vessel, Parr) 192 according to manufacturer instructions. Briefly, the pelletised biodried sample between 193 0,6-1g was electrically ignited in pure oxygen environment (30atm) and afterwards heat 194 of combustion produced, was monitored for subsequent calculations. Lower heating 195 value (LHV) was calculated from HHV by correcting it following the equation given by 196 Koppejan and Van Loo, (2012) and applied elsewhere (Gonzalez et al., 2019a). 197 198 2.4. Calculation of mass balances and performance indicators 199 Organic matter mineralization during biodrying was calculated according to the ash 200 conservation principle (Cai et al., 2012). Accordingly, final Volatile Solids (VS) weight 201 was calculated from VS content of representative products after homogenization and 202 grinding. The VS loss ratio was estimated for every stage (lag, thermophilic and late 203 mesophilic-cooling stages) from the percentage of cumulative O2 consumption 204 monitored in each stage. Then, those values were used to calculate moisture content 205 removal correcting it from monitored weight loss. Biodegradation of the bulking agent 206 was assumed to be negligible (Ponsá et al., 2011) as it was confirmed through dynamic 207 respirometry tests. 10 208 From the process efficiency point of view, daily drying rates are typically used to 209 assess experimental results. This parameter is clearly scale-dependent, and it does not 210 consider the organic carbon consumed. As the aim of the biodrying process is to obtain 211 a high-quality biomass fuel with high calorific potential, degradation of VS during the 212 process should be considered. Regarding this, the ratio of moisture removed per mass 213 unit of organic matter lost is presented as the appropriate indicator reflecting the 214 efficiency of the process, the so-called biodrying index (BI) (Hao et al., 2018). In the 215 current study, apart from the overall BI, the daily indices were also calculated to 216 identify and consider, stage per stage, the most important parameters affecting the 217 process. Moreover, for a more appropriate assessment of the process, energy 218 consumption and energy production potential parameters were introduced in the BI 219 calculation obtaining two new indices. Hence, those parameters could reflect the 220 energetic, economic and environmental viability of a certain biodrying process 221 performance. Consequently, the new Energetic Biodrying Index (EBI) and Biodrying 222 Performance Index (BPI) are presented (Equation 1 and 2, respectively). 223 1 1 𝐸𝐵𝐼 = · (𝐸𝑞𝑢𝑎𝑡𝑖𝑜𝑛 1) 𝑚 𝐸𝐶 𝑚 224 225 1 1 𝐵𝑃𝐼 = · 𝑥𝐸𝑃 (𝐸𝑞𝑢𝑎𝑡𝑖𝑜𝑛 2) 𝑚 𝐸𝐶 𝑚 226 227 Where mH2O is the water content lost (in mass) in the period of interest, mVS is the 228 VS content consumed in the same period, EC is the overall specific energy consumption 229 during the period (per dry mass of CS treated) and EP is the energy potential production 11 230 in terms of HHV of sieved product considering its specific production ratio (corrected 231 per dry mass of CS treated). 232 Additionally, an indicator referred to the mass conservation efficiency, that indirectly 233 measures the VS conservation capacity, is suggested by means of the specific 234 production ratio, as defined in Equation 3. 𝑚 𝑝𝑟𝑜𝑑𝑢𝑐𝑡 235 𝑆𝑝𝑒𝑐𝑖𝑓𝑖𝑐 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑟𝑎𝑡𝑖𝑜 = (𝐸𝑞𝑢𝑎𝑡𝑖𝑜𝑛 3) 𝑚 𝑓𝑒𝑑 236 Referring mTS product to the absolute mass of TS contained in the product and mTS-CS 237 fed to the absolute TS mass from CS fed in the beginning of the batch. 238 239 2.5. Gas and odour emissions: sampling and analysis 240 Samples were daily collected in Nalophan® bags by using a semi-spherical stainless- 241 steel flux chamber (Scentroid, IDES Canada Inc.) and a vacuum pump. CH4 and N2O 242 analysis were carried out using an Agilent 6890 N Gas Chromatograph (Agilent 243 Technologies, Inc.) equipped with a flame ionization detector and an electron capture 244 detector for CH4 and N2O detection, respectively. Total Volatile Organic Compounds 245 (tVOC), NH3 and H2S concentration in exhaust gas were measured in situ using a 246 MultiRAE Lite analyser (RAE Systems). The extended sampling method and gas 247 analysis can be found in González et al. (2019a). 248 249 2.6. Techno-economical assessment 12 250 An economic assessment for the implementation of the biodrying technology in 251 WWTPs was performed focusing solely in the CS valorisation step by using biodrying 252 process that produces a biomass fuel with economic value. 253 Real scale WWTP data were provided by Cirtec B.V. The specific CS production of 254 the WWTP studied was of 1.1E-02 t·PE-1·y-1, given total annual CS production of 2,920 255 tons while serving to 262,000 Population Equivalents (P.E.). Raw material 256 characteristics were defined as an average of the values obtained experimentally. From 257 this starting point, performance efficiencies experimentally obtained, were assumed for 258 the mass balance calculation. Real budget data were used for the calculation of 259 investment costs, assuming the construction of biodrying trenches made of concrete 260 including a cover for roofing and aeration system based on blowers. For yearly costs 261 calculation, energy consumption (electricity and diesel), personnel costs, B.A. cost, 262 pelleting, maintenance and insurance costs were estimated. The energetic consumption 263 of equipment was upscaled based on experimental data obtained and adapted to the 264 information provided by the industrial composting plant consulted (Aigües de Manresa 265 S.A., Spain). The market price value of end-products was determined according to the 266 specific energy content of biodried products and biomass energy selling price reported 267 by Avebiom (2019). Annual revenues were corrected from product selling earnings 268 considering yearly Operational Expendidures (OPEX). 269 The economical parameters calculated were: CAPEX, OPEX, Revenues, Net Present 270 value (NPV), Internal Rate of Return (IRR) and Payback Period Calculation detail of 271 NPV and IRR are facilitated in Equation 4 and 5. 272 𝑁𝑃𝑉 (€) = ∑ ( ) −𝐾 (𝐸𝑞𝑢𝑎𝑡𝑖𝑜𝑛 4) 13 (𝐵𝑡 − 𝐶𝑡 ) 273 =0 (𝐸𝑞𝑢𝑎𝑡𝑖𝑜𝑛 5) (1 + 𝐼𝑅𝑅)𝑡 274 where Bt are the annual benefits coming from the full-scale implementation of CS 275 biodrying system, in this specific case product selling and sludge management fees; Ct 276 are the annual costs of the implementation of the project (OPEX); r is the discount rate, 277 a value of 7% was used in this case, already used to assess this type of projects (Imeni et 278 al., 2019) and K are the investment costs expressed in € (CAPEX). T is the lifespan of 279 the project which was fixed to be of 25 years. 280 Payback period was calculated through the sum of annual cash flows over time until 281 a positive value is achieved, once this time is reached, net profit period would start. 282 From this initial framework, a breakeven point analysis was performed to find the 283 zero-profit scenario and determine the minimum feasible capacity of a treatment plant 284 (Imeni et al., 2019). 285 286 3. Results and discussion 287 3.1. Process evolution: temperature, moisture content and airflow rates 288 Temperature profiles, moisture and airflow evolution obtained during all the three 289 trials are shown in Fig. 1. In general, temperature profiles are comparable to those found 290 in literature regarding sewage sludge biodrying (Zhao et al., 2010). Maximum 291 temperatures achieved were equivalent for S1 and S3 (72ºC and 73.4ºC, respectively) 292 while temperature profiles were roughly similar until the process entered in a late 293 mesophilic stage, when the aeration rate clearly differed. As expected, the temperature 294 profile with S2 was different, reaching 55ºC after 24h and maintained for 2 more days. 14 295 After almost 4 days of operation, the maximum temperature peak of 63.5ºC was 296 achieved. Thermophilic temperatures were maintained for 5.3, 4.1 and 4.9 days with S1, 297 S2 and S3, respectively. Compared to what was obtained by other authors working on a 298 similar scale, both maximum temperatures achieved and the length of the thermophilic 299 stage with S1 and S3 were improved in the current study (Zhao et al., 2010; Vilegas and 300 Huiliñir, 2014). 301 As shown in Fig. 1, airflow rates supplied were significantly different for the three 302 strategies. Use of S2 high airflow rates (up to 3.5 L·min-1·kg-1 VS-CS) probably led to a 303 delayed temperature peak as well as a comparatively higher heat loss after temperature 304 peak (shorter thermophilic stage). However, thermophilic temperatures and satisfactory 305 biodrying performance were achieved, demonstrating that selected airflow rates for S2 306 were not high enough to impair biodrying process. As expected, Temperature Increasing 307 Phase (TIP) and Thermophilic stages in S1 followed the same trend than in S3. 308 However, the higher aeration rates used in S3 from day 6 on, significantly affected its 309 temperature profile. Probably, after day 8, the biogenic temperature generation was not 310 able to counterbalance the heat loss due to high aeration rate and consequently 311 temperature decreased to 25ºC and remained constant until the end of the experiment. 312 Accordingly, with S3, it could be assumed that only convective drying occurred after 313 day 8. 314 Considering the results shown in Table 2, maximum moisture removal ratio was 315 obtained when applying S2. MC removal ratios obtained were 55.0%, 62.4% and 57.5% 316 for S1, S2 and S3 respectively. When comparing these results but at a fixed VS mass 317 consumption, corresponding to the minimum value obtained among the 3 strategies that 318 was found for S3 (1.19 kg VS), the moisture removed applying S2 would still be 38% 15 319 and 11% higher than that removed with S1 and S3, respectively, demonstrating the high 320 efficiency of S2. 321 In terms of water removal efficiency, biodrying performance of CS was generally 322 high in comparison to the values reported in literature for similar low-porosity wastes, 323 where most of the MC removal ratio values found were between 45 and 60% (Zhao et 324 al., 2010; Huiliñir and Villegas, 2014). 325 Only co-biodrying of sludges with other biodegradable wastes is reported to improve 326 water removal efficiency (60-90%) (Zhang et al., 2018; Hao et al., 2018). 327 Due to the high temperatures achieved when applying S1 and S3 during the 328 thermophilic stage together with its longer duration, led to the most significant MC 329 removal occurring during this stage, compared to S2. Conversely, for S2, the MC 330 removal ratios were balanced among the thermophilic and mesophilic-cooling stages. 331 Cumulative oxygen consumption profiles were used to estimate VS consumption 332 ratios in each stage of the processes (Table 2). Considering B.A biodegradation 333 negligible, maximum VS consumption from cellulosic sludge (36.7% of initial VS-CS 334 content) occurred when applying aeration S1. In contrast, when applying S3, the lowest 335 VS consumption was determined (12.6% of the initial VS-CS content). Again, when 336 comparing the results at a fixed value of moisture removed (11 kg of water 337 corresponding to S3), there would be a 57% and 52% lower VS consumption for S2 and 338 S3, respectively, than in the case of S1. As expected, maximum VS biodegradation 339 occurred during the thermophilic stage, finding maximum absolute values in S1, more 340 than double than in the other two strategies. The low VS consumption obtained when 341 applying high airflow rates (along all stages of S2 and mesophilic-cooling stage of S3) 16 342 reinforces what other authors previously found about high airflow rates limiting 343 biological activity and degradation of organic matter (Huiliñir and Villegas, 2014). 344 345 It is worthwhile to highlight the potential effectiveness of all the tree strategies 346 implemented in terms of VS conservation, as even the highest VS consumption found 347 for S1 (14.9% of bulk mixture VS) resulted to be lower than most of the other authors 348 obtained (normally between 15 and 40% of VS consumption) (Zhao et al., 2010). 349 3.2 Process performance assessment 350 The most relevant performance assessment parameters to assess biodrying processes 351 are: (i) moisture removal; (ii) VS consumption and; (iii) the energy consumption along 352 the process. Moisture removal should be intended to be maximised, while VS 353 consumption must be limited, and energy consumption minimised. The Biodrying Index 354 (BI) is usually reported in the literature as performance efficiency index that interrelates 355 the first two of the mentioned key parameters. Additionally, Energetic Biodrying Index 356 (EBI) is presented in this study as a new index integrating the tree of them, adding an 357 energy consumption parameter into performance efficiency assessment. When the 358 above-mentioned indices are determined daily, they would allow the semi-continuous 359 process performance monitoring and the optimisation of the biodrying process 360 efficiency. 361 Process monitoring, by means of BI and EBI, for the three compared aeration 362 strategies are shown in Fig. 2a and b, respectively. When comparing the three strategies 363 assessed, the best final BI was obtained when applying S2 (9.8 kgH2O·kg-1VS), 364 followed closely by S3 (9.2 kgH2O·kg-1VS), and finally by S1 (4.3 kgH2O·kg-1VS). 365 Comparatively, the lowest BI obtained for S1 was indeed expected due to its higher VS 17 366 consumption, which doubled those found for S2 and S3. The limitation of organic 367 carbon mineralization is a key point for the improvement of biodrying performance 368 since it would affect the end-product quality as energy source. On the contrary, the best 369 BI obtained corresponds to S2 mainly due to the high MC removal ratio and the 370 moderate VS consumption. Accordingly, some authors also reported that the airflow 371 rate had more effect on moisture removal than in VS consumption (Vilegas and 372 Huiliñir, 2014). The comparison of CS biodrying performance results with values 373 reported in literature is presented in Table 3. Compared to other authors, all the 374 strategies tested, but especially S2 and S3, obtained satisfactory results in terms of 375 process efficiency, due to high MC removal ratios, but more particularly due to the 376 reduced VS consumption reported (Zhao et al., 2010; Villegas & Huiliñir, 2014). 377 However, some of the authors reported higher BI values (up to 20 kgH2O·kg-1VS) 378 compared to those presented in this study. This difference could be due to the 379 particularly low VS consumption associated to their low temperature profiles. In 380 addition, when comparing values reported in sludge co-biodrying studies, the use of co- 381 substrates resulted, in general, in higher MC removal ratios, but also significantly higher 382 VS consumption values, lowering in these cases the overall BI values (up to 6 383 kgH2O·kg-1VS) (Hao et al., 2018; Zhang et al., 2018; González et al., 2019a). 384 Some authors expose that overall water carrying capacity when using high airflow 385 rates should be substantially higher than that achieved due to high temperatures (Sharara 386 et al., 2012). The better drying performance of S2 compared to S3 during late 387 mesophilic stage is probably due to the difference in bulk temperature during that stage. 388 Although airflow rates are equivalent, the under mesophilic bulk temperatures found in 389 S3, clearly hampered the drying efficiency, compared to S2. The depletion of most 18 390 biodegradable VS during the first half of the S3 trial seemed to have reduced the 391 biogenic heat production in later stages, leading to low bulk temperatures. Thus, during 392 the late mesophilic stage, although high airflow rates can result in good MC removal 393 ratios, a minimum bulk temperature around 35-40ºC seems to be necessary for an 394 improved drying efficiency. 395 Fig. 2b shows the EBI profile along the three biodrying strategies assessed, 396 presenting clear differences among them, mainly related to their different aeration 397 strategies. 398 399 In overall terms, the most efficient strategy was found to be again S2 (0.99 400 kgH2O·kg-1VSkwh-1), followed closely by S3 (0.85 kgH2O·kg-1VS·kwh-1). 401 Conversely and although it was the one with the lowest overall energy consumption, 402 S1 obtained the lowest EBI value (0.62 kgH2O·kg-1VS·kwh-1), particularly due to the 403 high VS consumption, which did not actually improve moisture removal efficiency. 404 Energy consumption data in biodrying studies are scarce and only few studies 405 present some data. Sharara et al., (2012), determined energy consumption values around 406 1 kwh·kg-1mix, when treating livestock waste and using equivalent airflows than in S1. 407 Nevertheless, energy consumption data per water removed are more favourable, 0.4- 408 0.9 kwh·kg-1H2O in the present study vs. 2.2-2.5 kwh·kg-1H2O obtained in Sharara et 409 al., (2012), demonstrating the effective use of the biogenic heat produced combined 410 with appropriate aeration strategies to improve moisture removal. 411 In summary, when analysing the efficiency parameters proposed, S2 seems to be the 412 most efficient when considering moisture removal, BI and EBI. However, S3 also 413 shows promising results even considering that it can be further optimised during the 19 414 mesophilic-cooling stage. The information provided by the indices proposed together 415 with different aeration strategies used would certainly facilitate the upscaling of the 416 biodrying process. 417 418 3.3 Gaseous emissions 419 3.3.1 GHG emissions 420 Regarding Greenhouse Gases (GHG), maximum CH4 and N2O emission rates were 421 found during the first hours (0h for N2O and 24h for CH4) of the process. These 422 maximum emissions were probably related to anaerobic conditions during the 423 dewatering and shipping of raw materials (Han et al., 2018a). In fact, after adjusting the 424 initial structure, porosity and moisture content of the material, N2O stored in sludge was 425 probably stripped-out by forced aeration (Han et al., 2018a; González et al., 2019a) and 426 it was not detected anymore. CH4 emissions have been related to an inadequate mixture 427 structure and insufficient oxygen supply, leading to anaerobic conditions (Maulini- 428 Duran et al., 2013; Yuan et al., 2016). For instance, when applying S1, which can be 429 considered as the worst-case scenario, maximum daily emission rates found for N2O 430 and CH4 were 91.8 mg·d-1 and 16.3 mg·d-1, respectively. Additionally, the overall 431 emission factors calculated for N2O and CH4 were 6.8E-03 gN2O·kg-1TS and 2.6E-03 432 gCH4·kg-1TS, which were lower than values reported in biodrying and composting 433 literature (Han et al., 2018a; González et al., 2019a). Regarding the global warming 434 effect, the maximum cumulated value was 2.13 g CO2eq·kg-1TS, corresponding to S1. 435 This value is almost 3 times lower than the values reported in conventional sewage 436 sludge biodrying (González et al., 2019a) and even significantly lower than those of 437 sewage sludge composting (Yuan et al., 2016; Han et al., 2018a). 20 438 3.3.2 H2S, NH3 and total VOC emissions 439 In aerobic degradation processes such as composting or biodrying, H2S, NH3, and 440 tVOCs are the main compounds related to unpleasant odour emissions, which are 441 recognised to be a significant weakness of those processes (Han et al., 2018a). Emission 442 profiles of NH3 and tVOCs are shown in Fig. 3a and b, respectively. H2S was never 443 detected, reinforcing the effective aerobic conditions of biodrying mixtures with all the 444 aeration strategies implemented (Han et al., 2018b). NH3 and tVOC emissions followed 445 a typical profile where maximum NH3 emission peaks were related to thermophilic 446 temperatures whereas tVOCs were emitted mainly in the first days of operation 447 (Maulini-Duran et al., 2013; González et al., 2019a). Maximum emission rates for NH3 448 were detected with S1 (570 mg NH3·d-1), whereas peak emissions were 90% lower with 449 S2 (58.3 mg NH3·d-1), and 95.7% lower with S3 (25.8 mg NH3·d-1). Furthermore, the 450 highest overall NH3 emission factor was found during S1 (11.5E-01 g NH3·kg-1TS), 451 which emitted 80.9% and 96.1% more NH3 than strategies S2 (2.2E-02 g NH3·kg-1TS) 452 and S3 (4.5E-03 g NH3·kg-1TS) respectively. Comparatively, those values were always 453 lower than those of the sewage sludge biodrying (2.7E-01 g NH3·kg-1TS) (González et 454 al., 2019a) and composting processes (values found between 0.4 and 10.95 g NH3·kg- 1 455 TS) (Yuan et al., 2016; Han et al., 2018a). 456 Maximum tVOC emission rates found were 107.2; 41 and 80.8 mg C-VOC·d-1 for 457 strategies S1, S2 and S3, respectively. In all cases, those maximum values were detected 458 in the first hours (48h approximately), later decreasing to barely detectable values. 459 These results are in line with what other authors found during composting of sewage 460 sludge (Maulini-Duran et al., 2013, González et al., 2019b). The highest tVOC emission 461 factor was found when applying aeration S1 (1.4E-02 g C-VOC·kg-1TS), being 65.7% 21 462 and 30.7% higher than S2 (4.8E-03 g C-VOC·kg-1TS) and S3 (9.7E-03 g C-VOC·kg- 1 463 TS), respectively. Probably, the more adjusted aeration rates used in S1 and in the 464 thermophilic stage of S3, could lead to increase the anaerobic spots in the bulk mixture, 465 leading to significant tVOC emissions (Maulini-Duran et al., 2013). Compared to values 466 found in literature, there is limited information about tVOC emissions in biodrying 467 process and in this regard, only one work was found. All trials in the present study 468 emitted 55-85% less tVOCs than sewage sludge biodrying (3.1E-02 g C-VOC·kg-1TS) 469 (González et al., 2019a). 470 471 472 3.4 Quality assessment of final biodried products obtained 473 For a complete end-product quality assessment, both mixed and sieved end-products 474 were assessed in the present study and results are presented in Table 4. Although a 475 sustained combustion in a conventional biomass boiler can occur with a MC up to 55% 476 (Navaee-Ardeh et al., 2010), the maximum boiler efficiency is directly dependent on the 477 MC of the product, upgrading such efficiency up to 74-80% when reducing the MC 478 below 40% (Gebreegziabher et al., 2013). Besides, 20% of MC was claimed to be the 479 most appropriate value for pelleting process of Solid Recovered Fuels (SRF) (Rezaei et 480 al., 2020). All mixed products achieved MCs significantly lower than other authors 481 working with sludges or SRF (Shao et al., 2010; Cai et al., 2012; Villegas and Huiliñir, 482 2014; Yasar et al., 2018). 483 Apart from the MC, LHV is the other key parameter that determines the quality of 484 the biomass fuel produced. It seems that sustained combustion can occur from LHV 22 485 above 4MJ·kg-1 (Hao et al., 2018). All the biodried mixed products obtained in this 486 work presented LHV above 9MJ·kg-1, being equivalent to other conventional biomass 487 fuels used in boilers while the mixed product obtained when applying S3 can be 488 classified into group 4 according to the SRF quality standard (EN 15359). LHV 489 determined for all the three mixed end-products are, in general, higher than those found 490 in literature for conventional sewage sludge and pulp and paper mill sludge biodried 491 products (5.5-7.5MJ·kg-1 in the best cases) (Huiliñir and Villegas, 2014; Zhang et al., 492 2018). 493 When comparing to MSW biodried products the results are more variable. Some 494 authors obtained LHVs as high as 21MJ·kg-1 (Tambone et al., 2011), although such 495 high values can be related to their content of plastics and papers (Shao et al.,2010). 496 The mixed product quality assessment is the most usual study that is found in the 497 literature. Nevertheless, bulking materials (normally pruning waste or wood chips) are 498 hiding or diluting the real values corresponding to the waste streams that are being 499 valorised as biomass fuels, as it is the case in the present study (Table 4). Therefore, in 500 this study, the results corresponding to sieved materials will be prioritised. 501 502 Sieved materials presented always higher MC and consequently, lower LHV than 503 mixed materials. The lowest LHV of 5.4 MJ·kg-1 was found in the product obtained 504 when applying S1 and the highest value (7.9 MJ·kg-1) when applying S3. This last value 505 was comparable, to those obtained in other sewage sludge and paper mill sludge 506 biodrying studies but considering the mixed products obtained (Huiliñir and Villegas, 507 2014; Hao et al., 2018; González et al., 2019a). 23 508 Additionally, the energy production per energy consumed (EP/EC) and the biodrying 509 performance index (BPI) are presented in the current work as suitable indicators for the 510 evaluation of the process by means of end-product quality. Almost 2 and 3 kwh can be 511 recovered from sieved product per each kwh consumed in the process, demonstrating 512 the energetic efficiency of the process in all the three cases. Moreover, the new BPI 513 proposed in this work could be used as an overall biodrying process efficiency 514 indicator, facilitating the decision making and comparison of process efficiencies, as it 515 considers all the main factors involved in the biodrying performance, including the 516 quality, in terms of energy recovery potential of the end-products obtained. The best 517 BPI was achieved when applying S3 (35.1) mainly due to its high specific production 518 ratio. Comparatively, BPI values when applying S2 and S1 were 27% and 55% lower 519 than S3 values, respectively. 520 In general terms and considering all the efficiency indicators described in this work, 521 S3 was considered the best performing strategy. 522 Additionally, the end-product stability analysis was carried out indicating that these 523 materials were not totally stable (DRI above 3 g·kgVS-1·min-1 and AT4 above 200 g·kg- 1 524 VS). 525 Since S3 was considered the best control strategy, the techno-economic analysis 526 presented in section 3.5 was based on the results and data determined from this trial. 527 528 3.5 Techno-economical assessment 529 An economic model was developed and upscaled based on experimental results 530 obtained from S3, that it is considered the best performing strategy. In this study, only 531 CS valorisation step though biodrying was considered in the model as alternative 24 532 strategy to sludge disposal. The overall economic study of the WWTP after the 533 implementation of Cellvation® followed by the biomass fuel production through 534 biodrying would provide a more detailed analysis of the economic viability of this 535 technological innovation in a WWTP, however, this integrated assessment is out of the 536 scope of this current work. A breakeven point analysis of a hypothetic biodrying plant 537 was performed to find the minimum plant capacity size, in terms of population served, 538 which would lead to an economically sustainable scenario. To do so, economic 539 parameters of a biodrying plant were calculated according to variable mass flow of CS 540 treated, that directly depends on the treatment capacity of the WWTP related to PE 541 served. Table 5 specifies the main economic parameters and financial indicators of the 542 scenarios studied (more detailed information can be found in supplementary material, 543 Table 1S). For the smallest scale plants main OPEX and CAPEX costs were associated 544 to personnel costs and construction of windrows respectively. For largest scale plants, 545 main CAPEX costs were also related to construction of windrows while OPEX costs 546 were distributed among electricity, personnel and pelleting costs. In general terms, 53% 547 of the yearly revenues are related to product selling while the rest are due to avoided 548 costs from external sludge management or disposal. 25 549 550 * K refers to thousand 26 551 The zero-profit analysis determined that the minimum economically feasible WWTP 552 capacity is >60.000 PE. As an example, according to the Waterbase-UWWTD dataset 553 provided by EEA (EEA, 2020), the 56% of Spanish WWTP, providing services to 554 approximately 95% of total Spanish population, would have enough treatment capacity 555 to guarantee the economic viability of a complementary biodrying plant producing a 556 new source of renewable energy and significantly reducing the waste generated. 557 The IRR values obtained for WWTP providing services to more than 100.000 PE are 558 always above 40%, indicating the economical attractiveness of biodrying processes. 559 Complementarily, payback periods obtained for medium to large WWTP capacity, were 560 between 2 and 5 years, achieving worthy benefits (over yearly 100K €) in the case of 561 the largest plants (Table 3). 562 563 4. Conclusions 564 Two new process performance efficiency indexes, EBI and BPI, were proposed and 565 their relevance and appropriateness to monitor, assess and compare biodrying process 566 was confirmed. These two new indicators will contribute to better design, monitor and 567 assess current and future biodrying systems. 568 All three aeration strategies assessed (S1, S2 and S3) showed good performance 569 results and acceptable quality of the end-products obtained, compared to literature 570 results. Among them, S3 was selected as the best aeration strategy due to the highest 571 BPI values obtained and therefore the highest net energy recovery potential. Moreover 572 the three aeration strategies used showed low gaseous emissions and therefore low 573 environmental impacts are expected. Additionally, promising techno-economic 27 574 indicators were determined for the best aeration strategy (S3), obtaining IRR higher 575 than 40% and payback time of 2 years, for the best-case scenario (medium-large 576 WWTP). 577 In general terms, biodrying showed to be an adequate technology to valorise CS in 578 terms of economic and environmental indicators. 579 Acknowledgements 580 This study was carried out within the framework of the ‘SMARTPlant’ Innovation 581 Action, which has received funding from the European Union's Horizon 2020 research 582 and innovation program under grant agreement No. 690323. Joan Colón has received 583 funding from the 2018 call for Ramón y Cajal Grants from the Ministry of Science, 584 Innovation and Universities (reference RYC2018-026231-I) co-financed by the State 585 Research Agency and the European Social Fund. Authors thank CirTec B.V. for 586 providing CS samples for the study and Aigües de Manresa S.A for their availability 587 and orientation. Authors would like also thank Nicola Vitale (Univ. of Catania, Italy) 588 for his support developing the economic model. 589 590 References 591 APHA, American Public Health Association, 1995. Standard Methods for the 592 Examination of Water and Wastewater. https://www.standardmethods.org/ 593 Avebiom, Asociación Española de Valorización Energética de la Biomasa, 2019 Índice 594 de precios de la astilla de uso doméstico en España. Last update: June 2019. 28 595 Bayr, S., Rintala, J., 2012. Thermophilic anaerobic digestion of pulp and paper mill 596 primary sludge and co-digestion of primary and secondary sludge. Water Res. 46, 597 4713–4720. https://doi.org/10.1016/j.watres.2012.06.033 598 Cai, L., Gao, D., Chen, T. Bin, Liu, H.T., Zheng, G. Di, Yang, Q.W., 2012. Moisture 599 variation associated with water input and evaporation during sewage sludge bio-drying. 600 Bioresour. Technol., 117, 13–19. https://doi.org/10.1016/j.biortech.2012.03.092 601 Cellvation, B.V. 2018. Cellvation. Retrieved from https://www.cell-vation.com/ 602 Accessed date: November 2020 603 Crutchik, D., Frison, N., Eusebi, A.L., Fatone, F., 2018. Biorefinery of cellulosic 604 primary sludge towards targeted Short Chain Fatty Acids, phosphorus and methane 605 recovery. Water Res. 136, 112–119. https://doi.org/10.1016/j.watres.2018.02.047 606 CEN, E. (2011). 15359: 2011 Solid Recovered Fuels―Specifications and Classes. 607 Brussels, CEN. 608 Conca, V., da Ros, C., Valentino, F, Eusebi, L., Frison, N., Fatone, F., 2020. Long-term 609 validation of polyhydroxyalkanoates production potential from the sidestream of 610 municipal wastewater treatment plant at pilot scale. Chem. Eng. J. , 390, 124627. 611 https://doi.org/10.1016/j.cej.2020.124627 612 Da Ros, C., Conca, V., Eusebi, A.L., Frison, N., Fatone, F. 2020. Sieving of municipal 613 wastewater and recovery of bio-based volatile fatty acids at pilot scale. Water Res., 614 174.115633. https://doi.org/10.1016/j.watres.2020.115633 29 615 European Environment Agency (EEA), Waterbase - UWWTD: Urban Waste Water 616 Treatment Directive – reported data, v7, 2020, Permalink: 617 96f152fb61f04475a0f0bece739febb3. 618 Frei, K.M., Cameron, D., Stuart, P.R., 2004. Novel drying process using forced aeration 619 through a porous biomass matrix. Dry. Technol. 22, 1191–1215. https://doi.org/DOI 620 10.1081/drt-120038587 621 González, D., Guerra, N., Colón, J., Gabriel, D., Ponsá, S., Sánchez, A., 2019a. Filling 622 in sewage sludge biodrying gaps: Greenhouse gases, volatile organic compounds and 623 odour emissions. Bioresour. Technol., 291, 121857. 624 https://doi.org/10.1016/j.biortech.2019.121857 625 González, D., Colón, J., Gabriel, D., Sánchez, A., 2019b. The effect of the composting 626 time on the gaseous emissions and the compost stability in a full-scale sewage sludge 627 composting plant. Sci. Total Environ. , 654, 311–323. 628 https://doi.org/10.1016/j.scitotenv.2018.11.081 629 Gebreegziabher, T., Oyedun, A. O., Hui, C. W. 2013. Optimum biomass drying for 630 combustion–A modeling approach. Energy, 53, 67-73. 631 https://doi.org/10.1016/j.energy.2013.03.004 632 Han, Z., Sun, D., Wang, H., Li, R., Bao, Z., Qi, F., 2018a. Effects of ambient 633 temperature and aeration frequency on emissions of ammonia and greenhouse gases 634 from a sewage sludge aerobic composting plant. Bioresour. Technol., 270, 457–466. 635 https://doi.org/10.1016/j.biortech.2018.09.048 30 636 Hao, Z., Yang, B., Jahng, D., 2018. Spent coffee ground as a new bulking agent for 637 accelerated biodrying of dewatered sludge. Water Res., 138, 250–263. 638 https://doi.org/10.1016/j.watres.2018.03.049 639 Huiliñir, C., Villegas, M., 2014. Biodrying of pulp and paper secondary sludge: Kinetics 640 of volatile solids biodegradation. Bioresour. Technol., 157, 206–213. 641 https://doi.org/10.1016/j.biortech.2014.01.109 642 Imeni, S.M., Pelaz, L., Corchado-Lopo, C., Maria Busquets, A., Ponsá, S., Colón, J., 643 2019. Techno-economic assessment of anaerobic co-digestion of livestock manure and 644 cheese whey (Cow, Goat & Sheep) at small to medium dairy farms. Bioresour. 645 Technol., 291, 121872. https://doi.org/10.1016/j.biortech.2019.121872 646 Koppejan, J., Van Loo, S., 2012. The Handbook of Biomass Combustion and Co-firing, 647 The Handbook of Biomass Combustion and Co-firing. Routledge. 464 pages 648 https://doi.org/10.4324/9781849773041 649 Maulini-Duran, C., Artola, A., Font, X., Sánchez, A., 2013. A systematic study of the 650 gaseous emissions from biosolids composting: Raw sludge versus anaerobically 651 digested sludge. Bioresour. Technol., 147, 43–51. 652 https://doi.org/10.1016/j.biortech.2013.07.118 653 Navaee-Ardeh, S., Bertrand, F., Stuart, P.R., 2006. Emerging Biodrying Technology for 654 the Drying of Pulp and Paper Mixed Sludges. Dry. Technol. , 24 (7), 863–878. 655 https://doi.org/10.1080/07373930600734026 31 656 Larriba, O., Rovira-Cal, E., Juznic-Zonta, Z., Guisasola, A., Baeza, J.A. 2020. 657 Evaluation of the integration of P recovery, polyhydroxyalkanoate production and short 658 cut nitrogen removal in a mainstream wastewater treatment process. Water Res., 172, 659 115474. https://doi.org/10.1016/j.watres.2020.115474 660 Pagans, E., Barrena, R., Font, X., Sánchez, A., 2006. Ammonia emissions from the 661 composting of different organic wastes. Dependency on process temperature. 662 Chemosphere 62, 1534–1542. https://doi.org/10.1016/j.chemosphere.2005.06.044 663 Ponsá, S., Gea, T., Sánchez, A., 2010. Different Indices to Express Biodegradability in 664 Organic Solid Wastes. J. Environ. Qual., 39 (2), 706-712. 665 https://doi.org/10.2134/jeq2009.0294 666 Ponsá, S., Puyuelo, B., Gea, T., Sánchez, A., 2011. Modelling the aerobic degradation 667 of organic wastes based on slowly and rapidly degradable fractions. Waste 668 Manag., 31(7), 1472-1479. https://doi.org/10.1016/j.wasman.2011.02.013 669 Psaltis, P., Komilis, D.,2019. Environmental and economic assessment of the use of 670 biodrying before thermal treatment of municipal solid waste. Waste Manag., 83, 95- 671 103. https://doi.org/10.1016/j.wasman.2018.11.007 672 Ragazzi, M., Rada, E.C., Antolini, D., 2011. Material and energy recovery in integrated 673 waste management systems: An innovative approach for the characterization of the 674 gaseous emissions from residual MSW bio-drying. Waste Manag., 31 (9-10) 2085– 675 2091. https://doi.org/10.1016/j.wasman.2011.02.024 32 676 Raheem, A., Sikarwar, V. S., He, J., Dastyar, W., Dionysiou, D.D., Wang, W., Zhao, M. 677 2018. Opportunities and challenges in sustainable treatment and resource reuse of 678 sewage sludge: A review. Chem. Eng. J. . 337, 616-641. 679 https://doi.org/10.1016/j.cej.2017.12.149 680 Reijken, C., Giorgi, S., Hurkmans, C., Pérez, J., van Loosdrecht, M.C.M. 2018. 681 Incorporating the influent cellulose fraction in activated sludge modelling. Water Res.. 682 144, 104-111. https://doi.org/10.1016/j.watres.2018.07.013 683 Rezaei, H., Yazdanpanah, F., Lim, C. J., & Sokhansanj, S. 2020. Pelletization properties 684 of refuse-derived fuel-Effects of particle size and moisture content. Fuel Process. 685 Technol.. 205, 106437. https://doi.org/10.1016/j.fuproc.2020.106437 686 Richard, T.L., Hamelers, H.V.M.B., Veeken, A., Silva, T., 2002. Moisture relationships 687 in Composting Processes. Compost Sci. Util. 10 (4), 286–302. 688 https://doi.org/10.1080/1065657X.2002.10702093 689 Rihani, M., Malamis, D., Bihaoui, B., Etahiri, S., Loizidou, M., Assobhei, O., 2010. In- 690 vessel treatment of urban primary sludge by aerobic composting. Bioresour. Technol. 691 101, 5988–5995. https://doi.org/10.1016/j.biortech.2010.03.007 692 Sadaka, S., Ahn, H., 2012. Evaluation of a biodrying process for beef, swine, and 693 poultry manures mixed separately with corn stover. Appl. Eng. Agric. 28, 457–463. 694 https://doi.org/10.13031/2013.41482 33 695 Shao, L.M., Ma, Z.H., Zhang, H., Zhang, D.Q., He, P.J., 2010. Bio-drying and size 696 sorting of municipal solid waste with high water content for improving energy recovery. 697 Waste Manag.. 30 (7), 1165–1170. https://doi.org/10.1016/j.wasman.2010.01.011 698 Sharara, M.A., Sadaka, S., Costello, T.A., Vandevender, K., 2012. Influence of aeration 699 rate on the physio-chemical characteristics of biodried dairy manure-wheat straw 700 mixture. Appl. Eng. Agric., 28 (3), 407–415. 701 Tambone, F., Scaglia, B., Scotti, S., Adani, F., 2011. Effects of biodrying process on 702 municipal solid waste properties. Bioresour. Technol., 102, 7443–7450. 703 https://doi.org/10.1016/j.biortech.2011.05.010 704 Toledo, M., Márquez, P., Siles, J.A., Chica, A.F., Martín, M.A., 2019. Co-composting 705 of sewage sludge and eggplant waste at full scale: Feasibility study to valorize eggplant 706 waste and minimize the odoriferous impact of sewage sludge. J. Environ. Manage. 247, 707 205–213. https://doi.org/10.1016/j.jenvman.2019.06.076 708 US Department of Agriculture and US Composting Council (2001). Test Methods for 709 the Examination of Composting and Compost. Edaphos International, Houston. 710 Villegas, M., Huiliñir, C., 2014. Biodrying of sewage sludge: Kinetics of volatile solids 711 degradation under different initial moisture contents and air-flow rates. Bioresour. 712 Technol., 174, 33–41. https://doi.org/10.1016/j.biortech.2014.09.136 713 Winkler, M.K.H., Bennenbroek, M.H., Horstink, F.H., van Loosdrecht, M.C.M., van de 714 Pol, G.J., 2013. The biodrying concept: An innovative technology creating energy from 34 715 sewage sludge. Bioresour. Technol. 147, 124–129. 716 https://doi.org/10.1016/j.biortech.2013.07.138 717 Yasar, A., Shabbir, S. A., Tabinda, A. B., Nazar, M., Rasheed, R., Malik, A., Mukhtar, 718 S. 2018. Refuse-derived fuels as a renewable energy source in comparison to coal, rice 719 husk, and sugarcane bagasse. Energy Sources, Part A Recover. Util. Environ. Eff., 720 41(5), 564-572. https://doi.org/10.1080/15567036.2018.1520340 721 Yuan, J., Chadwick, D., Zhang, D., Li, G., Chen, S., Luo, W., Du, L., He, S., Peng, S., 722 2016. Effects of aeration rate on maturity and gaseous emissions during sewage sludge 723 composting. Waste Manag., 56, 403–410. 724 https://doi.org/10.1016/j.wasman.2016.07.017 725 Zhang, X., Yan, S., Tyagi, R.D., Surampalli, R.Y., Valéro, J.R., 2014. Wastewater 726 sludge as raw material for microbial oils production. Appl. Energy 135, 192–201. 727 https://doi.org/10.1016/j.apenergy.2014.08.078 728 Zhang, D., Luo, W., Yuan, J., Li, G., 2018. Co-biodrying of sewage sludge and organic 729 fraction of municipal solid waste: Role of mixing proportions. Waste Manag., 77, 333– 730 340. https://doi.org/10.1016/j.wasman.2018.04.016 731 Zhao, L., Gu, W.M., He, P.J., Shao, L.M., 2010. Effect of air-flow rate and turning 732 frequency on bio-drying of dewatered sludge. Water Res., 44 (20), 6144–6152. 733 https://doi.org/10.1016/j.watres.2010.07.002 734 735 35 736 DRI Total Volatile HHV LHV cE N-TKN N-NH4 (gO2 Type of sludge Solids Solids (MJ kg- (MJ pH (mS (%, db) (%, db) 1 kgVS- (%, wb) (%, db) TS) kg-1) cm-1) 1h-1) Cellulosic 25-37 85-93 sludge 3-12 1.9-2.5 18-19 2.1-4.9 4.7-6.9 0.5-1.6 2.3-3 Primary sludge 5-28 60-80 1.5-4 5.6-6.9 Secondary sludge 15-25 52-76 3-6 11-17 0.5-0.9 6.4-7.9 3-7 Mixed sludge 26-38 60-70 2.5-4 0.5-1 5.9-7.1 1.2-1.8 6-7 Anaerobically digested sludge 17-38 53-70 2.6-7 0.7 7.6-7.9 1.2-2.1 1.2-3.7 Pulp & Paper mill sludge 19-26 80-85 0.5-5 18-21 6.2-7.8 737 738 739 740 741 742 743 744 Table 1. Main physic-chemical characteristics of CS and other conventional sludges 745 Data gathered from: Navaee-Ardeh et al., 2006; Pagans et al., 2006; Rihani et al., 2010; 746 Bayr & Rintala, 2012; Maulini-Duran et al., 2013; Zhang et al., 2014; Crutchik et al., 747 2017; Hao et al., 2018; Zhang et al., 2018; Toledo et al., 2019; Da Ros et al., 2020. 36 748 749 750 751 752 753 754 37 755 Table 2. Air supplied and overall mass balances in the different stages of cellulosic sludge biodrying trials operated with different control 756 strategies. Time is given in days (d) and air supply as an average of the period in m3 per kg of VS fed from CS. Duration Air supply Weight loss Water removal VS consumption Days Total m3 Av. m3kg-1VS-CS·d-1 Kg (%) Kg (%) Kg (%) S1 TOTAL 12.2 128.5 1.4 14.6 11.8 (55.0%) 2.8 (14.9%) TIP 0.6 1.4 0.3 1.2 (8.2%) 1.1 (5.1%) 0.1 (0.6%) THERMOPHILIC ST. 5.4 76.6 1.9 10.4 (71.2%) 8.4 (39.0%) 2.0 (10.8%) MESOPHILIC ST. 6.3 50.7 1.1 3.0 (20.6%) 2.3 (10.9%) 0.7 (3.5) S2 TOTAL 13.0 207.4 2.7 13.6 12.3 (62.4%) 1.26 (10.0%) TIP 1.0 2.2 0.4 0.4 (2.9%) 0.4 (1.8%) 0.04 (0.3%) THERMOPHILIC ST. 4.1 101.8 4.2 7.3 (53.7%) 6.8 (34.2%) 0.5 (4.2%) MESOPHILIC ST. 8.0 103.4 2.2 5.9 (43.4%) 5.2 (26.3%) 0.7 (5.4%) S3 TOTAL 13.2 211.6 1.7 12.2 11.0 (57.5%) 1.19 (6.9%) TIP 1.1 2.5 0.2 0.2 (1.6%) 0.2 (0.8%) 0.05 (0.3%) THERMOPHILIC ST. 4.9 64.6 1.4 8.8 (72.1%) 8.0 (41.6%) 0.8 (4.8%) MESOPHILIC ST. 7.2 144.4 2.1 3.2 (26.2%) 2.9 (15.1%) 0.3 (1.8%) 757 *The mass balances were done according to bulk mixtures, to be consistent with other authors. **TIP is referred to Temperature Increasing 758 Phase 38 759 Table 3 Comparative biodrying overall efficiencies in literature for similar high moisture organic wastes. MC VS Co- Specific Initial Final removal consumption BI EBI Raw substrate aeration MC MC Reference Scale ratio ratio material (Y/N; (L ·min- kg H2Or· kg H2Or· kgVSc- 1 (%) (%) (%) (%) which) ·kgVS-1) kgVSc-1 1 ·Kwh-1 S1 51.9 35.1 55.0 14.9 4.3 0.62 Cellulosi Bench This study N S2 57.8 32.5 62.4 10.0 9.8 0.99 c sludge (100L) S3 51.8 31.5 57.5 6.9 9.2 0.85 Y González et Secondar (diatomac Bench Variable 54.6 35.9 58.8 14.5 5.7* al., 2019a y sludge eous (100L) earth) Dewatere Y (Spent Hao et al., Lab 68.3- d sewage Coffee 1.37 46.2 79.7 43.5 4.37 2018 (28.3L) 71.6 sludge Ground) Dewatere Zhang et Y Lab 45.1- d sewage 0.49-0.56 70 45.1-78.6 35.1-46.7 3.3-4.6 al., 2018 (MSW) (19.44L) 68.3 sludge Villegas Dewatere and d bench 51- N 1.05-3.14 58 16.9-24 5-14.3 16-20 Huiliñir, secondar (64L) 52.5 2014 y sludge 760 39 761 Table 3 cont. Pulp and Paper Huiliñir and secondary N Lab (9L) 0.51-5.26 64.4-65.2 62-45 20-58 0-18 2.5-12.7* Villegas, 2014 sludge Winkler et al., Dewatered Industrial N Variable 75 27.4 90.5 26 11.1* 183.6* 2013 sewage sludge (1900 m3) Cai et al., 2012 Sewage sludge N Pilot (1.6m3) Variable 66.1 54.7 46.1 Sharara et al., 2012 Dairy manure N Bench (147 L) 0.05-1.5 55.9 28-35 70.7-79.1 26.3-41.9 2.6-3.2* 24.7-346.7* Beef manure 59 30 59 8.1 15.5* 0.126* Sadaka & Ahn, Swine manure N Pilot (0.9 m3) 0.65 60 28 58 5.8 19.8* 0.08* 2012 Poultry manure 61 40 53 5.9 19.0* 0.11* Tambone et al., residual MSW N Industrial Variable 32.7 17.8 65.5 29 2.26* 2011 Shao et al., 2010 MSW N Bench (150L) 1.4 73 48.3 79.9 37.3 7.02* Dewatered Zhao et al., 2010 N Bench (81L) 3.1-6.1 67.8 30.5-41.9 57.5-68.2 31.0-36.7 5.9-6.1* sewage sludge Pulp and Paper Frei et al., 2004 N Pilot (1m3) 52.5-75.5 34.3-59.5 47-53.5 5.5-18 5.9-21.7* mixed sludge 762 *Estimated from the values provided in the work 40 763 Table 4. Quality assessment parameters of cellulosic sludge biodrying end-products S1 S2 S3 Parameter Sieved Mixture Sieved Mixture Sieved Mixture MC (%, w.b.) 57.3 35.1 51.4 35.5 43.3 31.5 VS (%, d.b.) 88.7 88.7 85.5 84.7 91.9 94.2 17.1 ± 17.2 ± 16.9 ± 16.9 ± 17.71 ± -1 HHV (MJ·kg TS) 17.1 ± 0.05 0.1 0.1 0.3 0.1 0.00 HHV % lost from 9.9 4.2 4.9 3.4 10 6.1 initial 9.5 ± 6.57 ± 9.4 ± 7.88 ± 10.6 ± LHV (MJ·kg-1) 5.4 ± 0.03 0.2 0.06 0.2 0.07 0.00 LHV % gained from 46.1 27.0 206.9 53.5 60.8 30.5 initial Specific production ratio (kgTS product·kg- 0.65 - 0.81 - 0.87 - 1 TS-CS fed EP/EC (kwh·kwh-1) 1.8 - 2.1 - 3.1 - BPI 15.7 - 25.6 - 35.1 - 764 765 766 767 768 41 769 Table 5. Economic parameters and financial indicators of variable CS input scenarios. NPV values are given considering a lifetime of 25 770 years and considering a discount rate of 7%. 20K 40K 60K 80K 100K 150K 200K 250K 300K 400K 500K 750K 1000K 20.012 23.878 27.743 33.309 37.664 66.762 99.082 136.888 140.368 176.319 246.721 351.540 467.793 CAPEX (€) € € € € € € € € € € € € € 28.393 32.105 35.663 39.740 43.673 58.595 67.509 82.288 99.608 122.881 148.187 234.227 339.788 OPEX (€·y-1) € € € € € € € € € € € € € REVENUE 11.936 23.872 35.808 47.743 59.679 89.519 119.359 149.198 179.038 238.717 298.397 447.595 596.794 (€·y-1) € € € € € € € € € € € € € BENEFITS -16.457 -8.233 8.004 16.006 30.924 51.849 66.910 79.430 115.837 150.210 213.368 257.006 145 € (€·y-1) € € € € € € € € € € € € - - - 143.32 323.69 645.37 1.071.6 1.388.0 1.674.8 2.489.4 3.194.1 4.567.2 5.440.2 NPV (€) 410.13 221.84 29.968 0€ 8€ 7€ 96 € 17 € 11 € 90 € 70 € 39 € 84 € 4€ 6€ € IRR (%) - - - 23% 42% 46% 52% 49% 56% 66% 61% 61% 55% PAYBACK INF INF INF 5 3 3 3 3 2 2 2 2 2 (y) 771 772 773 42 774 Fig. 1 Temperature, airflow rate and MC profiles during experimental trials 775 implementing strategies S1 , S2 and S3 are shown in figures a, b and c, respectively. 776 Arrows indicate whenever mixture was turned. 43 777 778 Fig. 2 Daily comparative biodrying performance efficiency indexes: (a) Biodrying 779 Index (kgH2O kgVS-1) and (b) Energetic Biodrying Index (kgH2O kgVS-1 kwh-1), the 780 grey area indicates roughly thermophilic stages during trials. Axes for thermophilic and 781 the rest of the process of EBI profiles differ as they were adjusted to the values obtained 782 in each phase. 44 783 784 Fig. 3. NH3 and tVOC emission patterns during the three biodrying aeration strategies 785 implemented. 45 786 46