Assessment of Energy-Positive Wastewater Treatment System Design of Different Industrial Wastewater Streams


 Activated sludge (AS) process has been used as the conventional industrial wastewater treatment process in the past decades. However, intensive aeration requirements made it impossible for WWTPs to be energy positive. In addition, there are few cases of research assessing the efficiency and economic feasibility of innovative technologies in treating industrial wastewater to achieve energy positive. This study tries to assess the effect of different kinds of industrial wastewater on treatment efficiency, unit energy, environmental sustainability, and unit cost of treatment systems using innovative technologies such as micro sieving, UASB, and PN/A. Our Excel model showed that micro-sieving could remove 32.50-39.72% of COD from industrial wastewater. UASB reactor could removes 15.8%-53.5%, 14.0%-49.0% and 22.9%-51.0% of COD from three different wastewater streams. Mean unit energy production by the innovative system could reach 1.80, 1.77, and 1.73 kWh/kg BODremoved, respectively. The average unit cost of three kinds of wastewater is 0.54, 0.57, and 1.12 $/kg CODremoved, respectively. Treating meat processing wastewater with innovative technologies is the economical treatment method. However, this method could not be considered as the sustainable design due to the high effluent COD concentration. The economic advantages and limitations of innovative technologies in treating industrial wastewater have been quantified by energy, cost and environmental indicators. These metrics provide valuable references for future sustainable design of industrial wastewater treatment systems.


41
Industrial wastewater has vastly different water quality. For example, the water 42 quality of food processing, pulp and paper, textile, chemical, pharmaceutical, 43 petroleum, tannery, and manufacturing industries varied significantly [1]. Major 44 wastewater quality parameters include chemical oxygen demand (COD), biochemical 45 oxygen demand (BOD), suspended solids (SS), ammonium nitrogen (NH4 + _N), heavy 46 metals, pH, color, turbidity, and biological parameters. Compared with municipal 47 wastewater, industrial wastewaters usually have a high organic matter concentration 48 and extreme physicochemical characters (e.g., pH, temperature, salinity), and humic 49 substances that may inhibit biological treatment processes. In addition, municipal 50 wastewater has a low strength concentration of COD (250-800 mg/L), whereas strong 51 (>1,000 mg COD/L) to extremely strong wastewaters is often produced by industries 52 [2]. Olive mills and beverage production industries could generate extremely strong 53 industrial wastewaters (COD>200,000 mg/L) [3,4]. Characteristics of industrial 54 wastewaters strongly depend on the type of industrial wastewaters and industrial 55 processes. Water used by meat processing industries accounts for 29% of the 56 agricultural freshwater worldwide [5,6]. Food processing wastewater is typically 57 generated from the slaughtering houses. Typically, there is a large quantity of 58  T=temperature=25℃ 183 fns = non-biodegradable, soluble influent COD fraction (g COD/g COD) 184 fnp = non-biodegradable particulate influent COD fraction (g COD/g COD) 185 Rsu = anaerobic sludge age in the UASB reactor Yan = yield coefficient in an anaerobic environment (0.05 g VSS/g COD) as 187 determined by Cavalcanti et al. [ To compare the cost of treatment systems, the total capital costs ($) were annualized 306 over the expected lifetime of a WWTP. Economic lifetime and interest rate were 307 assumed to be 20 years and 6% [76]. Capital costs ($) were annualized capital costs 308 ($/yr) based on the interest rate and lifetime. The total unit cost was calculated as the 309 summation of the unit capital and the unit O&M costs. 310 industrial wastewater. It shows that the TSS/COD ratio of textile wastewater is higher 319 than those of other industrial wastewaters. Due to relatively high TSS concentration, 320 micro sieving shows TSS removal of greater than 50% and significantly higher COD 321 removal than the primary clarifier, as shown in Figure 4. For tannery wastewater, meat 322 processing wastewater, and textile wastewater, the average COD removal by micro 323 sieving is 35.36%, 39.08%, and 32.50%, respectively, which are higher than that by 324 primary clarifier. Textile wastewater has a relatively lower TSS concentration than 325 other industrial wastewaters in Table 1. However, primary clarifier in textile 326 wastewater treatment shows higher COD removal than tannery wastewaters due to the 327 high TSS/COD ratio of textile wastewater, as shown in  Table 3 shows the Pearson's correlations between COD removal of two systems, 336 the concentration of TSS and TSS/COD. Compared with TSS concentration, the ratio 337 between TSS and COD showed stronger positive correlations with COD removal at the 338 0.01 level. Therefore, regression models of COD removal are developed based on 339 TSS/COD ratio. The trend lines for predicting COD removal of primary treatment from 340 TSS/COD are depicted in Figure 5. TSS concentration of textile wastewater is lower 341 than those of other industrial wastewaters. However, textile wastewater has a relatively 342 low concentration of COD, as shown in Table 1. The particulate COD fraction of textile 343 wastewater is higher than those of other industrial wastewaters. Since TSS/COD ratio 344 is associated with particulate COD fraction, COD removal of two systems rises with 345 the increasing TSS/COD ratio of influent wastewater as shown in Figure 5. Therefore, 346 TSS/COD ratio is the key parameter for determining COD removal of primary 347 treatment. Due to the high particulate matter fraction, primary treatment shows 348 excellent separation performance during treatment of textile wastewater. 349 Table 3 Correlations between COD removal, TSS concentration, and TSS/COD ratio The higher the BOD/COD ratio, the higher is the 358 biodegradability. Biodegradability of meat processing wastewater, tannery wastewater, 359 and textile wastewater are shown in Figure 3 which illustrates the average BOD/COD 360 ratio of 0.412, 0.327, and 0.314, respectively. Therefore, meat processing wastewater 361 has higher biodegradability than other industrial wastewaters. Figure 6 shows the 362 BOD/COD ratio of untreated wastewater and treated wastewater by primary clarifier 363 and micro-sieving. Biodegradability rises with the increasing BOD/COD ratio of 364 untreated wastewater. Figure 6 demonstrates that the slope of the linear fitting equation 365 for system B is greater than that of system A, which illustrates that micro-sieving could 366 improve the biodegradability of wastewater more significantly as compared with the 367 primary clarifier. Primary treatment increases the biodegradability of wastewater by 368 reducing the particulate non-biodegradable COD fraction of industrial wastewater. As 369 compared to system A, the BOD/COD ratio of wastewater increased by 0.59%-7.50%, 370 0.60%-3.57%, and 0.01%-2.19% for meat processing wastewater, tannery wastewater, 371 and textile wastewater, respectively. Due to the relatively low TSS concentration of 372 textile wastewater, TSS removal by primary clarifier and by micro sieving and variation 373 of BOD/COD ratio is smaller. Meat processing wastewater has higher biodegradability 374 and TSS concentration, micro sieving increases mean and maximum BOD/COD ratio 375 to 0.48 and 0.63. Hence, the biodegradability of industrial wastewater with a high 376 concentration of TSS is easier to be improved by micro-sieving. Due to high COD removal of micro sieving and UASB reactor in system B, COD 381 removal of aerobic treatment in system B decreases by 40.7%-66.5%, 33.6-60.0%, and 382 23.0%-60.0% for meat processing wastewater, tannery wastewater, and textile wastewater as compared to aerobic treatment in system A in Figure 7. Micro-sieving 384 and UASB could reduce oxygen demand for all kinds of industrial wastewater. In 385 secondary treatment, the UASB reactor removes 15.8%-53.5%, 14.0%-49.0%, and 386 22.9%-51.1% of COD for three kinds of industrial wastewater. Because of the higher 387 BOD/COD ratio of treated meat processing wastewater by micro-sieving, UASB could 388 convert more COD to CH4 than other industrial wastewater, which resulted in less 389 oxygen demand to remove the remaining COD.  does not meet the current discharge standards. This is due to a much high concentration 417 of COD, relatively low biodegradability (BOD/COD ratio), and the relatively high 418 content of refractory soluble organic matter. Most of the remaining COD, which is the 419 soluble non-biodegradable matter, is not easily removed by primary physical treatment 420 and biological processes. Physical-chemical treatment such as adsorption might be 421 required to remove soluble substances by the accumulation of those substances on 422 activated carbon to increase COD removal efficiency and meet the discharge 423 requirement. Figure 9 illustrates that the effluent of some meat processing wastewater 424 and textile wastewater achieve the discharge requirement of several countries due to 425 their low concentration of COD and relatively high BOD/COD ratio. Therefore, it is 426 necessary to study the relationship between the total COD removal of primary treatment 427 and biological processes and wastewater quality. The squared correlation coefficient R 2 represents that the BOD/COD ratio and 442 TSS/COD ratio of influent wastewater can explain 92.9% of COD removal. The 443 BOD/COD ratio and TSS/COD ratio are positively related to COD removal efficiency. 444

Results and Discussions
Micro-sieving and UASB could achieve higher COD removal when treating industrial 445 wastewater with higher particulate matter fraction and biodegradability. Effluent COD 446 concentration could be calculated based on COD removal and influent COD 447 concentration. Therefore, characteristics of wastewater could be used for predicting 448 effluent COD concentration of industrial wastewater in a system with primary treatment 449 and biological processes to determine if physical-chemical treatment is required. 450 Table 5 Correlations between COD removal, BOD/COD ratio and TSS/COD ratio of 451 influent wastewater 452

Energy demand and production 454
3.3.1 Theoretical and actual energy production of UASB In this study, empiric equations (1-5) were used to estimate theoretical COD 456 removal and unit energy consumption of UASB, as shown in Table 6. Theoretical COD 457 removal, unit methane production, and unit energy production at 25 ℃ are 54%-88%, 458 0.15-0.30 m 3 CH4/kg COD removal, and 0.52-1.03 kWh/kg COD. 459  Table 7 lists the data of COD removal and methane production of UASB in treating 466 food processing wastewater. These data were collected from published peer-reviewed 467 papers [109][110][111]. Actual COD removal, unit methane production, and unit energy 468 production at 25 ℃ are 43%-95%, 0.01-0.38 m 3 CH4/kg COD removal, and 0.03-1.30 469 kWh/kg COD. As shown in Table 6 Table 8 lists the actual unit energy consumption of PN/A is 0.8-1.92 kWh/kg N 480 removed. Based on equations (9-10), 1.32kWh/kg NH4 + _N was used as the theoretical 481 unit energy consumption of PN/A, which is close to the actual value, as shown in Figure  482 11. This proves the validity of the theoretical unit energy consumption. According to 483 the benchmarking method reported by Yang Figure 12 shows that 496 electricity consumption rises as the influent concentration of NH4 + _N and soluble 497 biodegradable organic nitrogen (sbON) increases. Therefore, unit energy consumption 498 (kWh/kg N removal) is used for comparing energy efficiency of different industrial 499 wastewater systems as shown in Figure 13. The oxygen requirement of PN/A process 500 and nitrification process are 1.83 and 4.57 g O2/g NH4_N converted. PN/A process in 501 system B effectively reduces aeration consumption for removing NH4_N. Therefore, 502 system B shows lower unit energy consumption than system A. In Figure 13, mean 503 unit energy consumptions of system B for meat processing wastewater, tannery 504 wastewater, and textile wastewater are 1.49, 1.37, and 1.39 kWh/kg N removal, 505 respectively. Since primary treatment and UASB removes COD without using oxygen, 506 most of O2 is used for the PN/A process, especially in system B. Therefore, mean unit 507 energy consumptions of system B are close to than actual unit energy consumption of 508 PN/A, as shown in Figure 11. Except for BOD that is converted to CO2 in aerobic treatment, most of soluble BOD 514 (sBOD) in industrial wastewater and biodegradable particulate BOD (bpBOD) in 515 primary and secondary sludge are converted to CH4 to produce electricity and heat by 516 UASB and side stream anaerobic digester. Figure 14 illustrates energy production rises 517 with increasing influent BOD concentration. The unit energy production (kWh/kg 518 BOD removal) is used as an indicator in assessing the influence of different industrial 519 wastewater on two treatment systems, as shown in Figure 15. The average unit energy 520 production of system B for meat processing wastewater, tannery wastewater, and textile 521 wastewater are 1.80, 1.77, and 1.73 kWh/kg BOD removal, respectively. The unit 522 energy production for food processing wastewater is higher than that of other industrial 523 wastewaters. Meat processing wastewater has higher BOD concentration and 524 biodegradability (BOD/COD ratio), as shown in Table 1 and Figure 3. This increases 525 biogas production efficiency and unit energy production of UASB reactors, as shown 526 in Figure 16. In addition, sludge production from the PN/A process and denitrification 527 process increase CH4 production of side stream anaerobic digester. Therefore, the 528 energy positive system for treating meat processing wastewater could produce more 529 energy with the same BOD removal than that for treating tannery wastewater and textile 530 wastewater.  Figure 17 shows the influence of different industrial wastewater on energy recovery 540 of two treatment systems. System A for treating industrial wastewater could not 541 produce enough electricity to satisfy the energy demand for oxygen production due to 542 the energy ratio of less than 1. Since UASB effectively converts COD to CH4 and PN/A 543 decreases energy consumption for removing N, innovative technologies could help 544 industrial wastewater treatment achieve electrical self-sufficiency, as shown in Figure  545 17. Meat processing wastewater has higher BOD concentration and BOD/COD ratio. 546 However, N concentration of meat processing wastewater is higher than other industrial 547 wastewaters. Energy consumption and energy production are influenced by BOD and 548 N concentration, respectively. In system B, energy recovery ratio is positive related to 549 BOD/N ratio, as shown in Figure 18. High BOD and N concentration of meat 550 processing wastewater leads to relatively lower BOD/COD ratio and energy recovery 551 ratio. Although textile wastewater has lower BOD concentration, the value of the 552 energy ratio for textile treatment system B could be as high as 17.7 because of lower N 553 concentration and higher BOD/N ratio. The minimal energy ratio is 7.0 which is seven 554 times higher than 1. The mean energy ratio is 14.3. Therefore, BOD/N ratio 555 significantly affects energy recovery of treatment system with innovative technologies.

Environmental indicators 562
The contributors of system A and B to global warming and eutrophication potential 563 are shown in Figure 19 and 20. Since the global warming potential is higher than 0, 564 activated sludge process with nitrification-denitrification for treating industrial 565 wastewater could not be considered as the sustainable design. Energy significantly 566 affects global warming potential of treatment systems. System B could produce enough 567 electricity to satisfy the aeration demand. The energy production could offset the 568 overall global warming potential in system B, as shown in Figure 19. Therefore, 569 achieving energy positive by innovative technologies could effectively reduce impacts 570 of industrial wastewater treatment on the global warming. Due to high energy recovery 571 ratio, the eutrophication potential of treatment system B is lower than that of system A, 572 as shown in Figure 20. Since the eutrophication potential is affected by effluent 573 pollutant mass, meat processing wastewater treatment has higher eutrophication 574 potential than textile wastewater treatment. Many researchers did not consider the 575 physical-chemical treatment when designing wastewater treatment system [76, 77, 576 116]. The physical-chemical treatment is not necessary because non-biodegradable 577 COD of municipal wastewater is low enough to meet discharge requirements. 578 However, meat processing wastewater has high concentration of COD and non-579 biodegradable COD. This results in a high eutrophication potential of food processing 580 wastewater treatment. UASB and activated sludge process could not remove non-581 biodegradable matter. Therefore, the physical-chemical treatment is required to be 582 applied in the future food processing wastewater treatment design for removing non-583 biodegradable COD. 584

Cost analysis 591
Unit cost analysis of treatment systems based on different industrial wastewater 592 quality compares the cost of these systems. Since COD removal by primary treatment 593 and secondary treatment are positively related to TSS/COD ratio and BOD/COD ratio, 594 unit cost metric ($/kg CODremoved) is used as the indicator in this study. Figure 21  595 compares the unit cost ($/kg CODremoved) of two systems for treating different industrial 596 wastewater. Although application of UASB and PN/A in system B increase the 597 corresponding cost, UASB and PN/A reduces influent BOD and N loading and required 598 size of activated sludge process with nitrification-denitrification. Figure 21  599 demonstrates that the unit cost of system B is lower than that of system A. Since 600 economic income due to energy recovery and cost saving is higher than the increased 601 cost of innovative technologies, innovative biological technologies could replace the 602 old process. In system B, the average unit cost for meat processing wastewater, tannery 603 wastewater, and textile wastewater are 0.54, 0.57, and 1.12 $/kg COD removal, 604 respectively. The unit cost for meat processing wastewater is lower than that for the 605 other wastewaters. Compared with other wastewater, meat processing wastewater has 606 a higher biodegradability and concentration of BOD, which could produce a large 607 amount of CH4 for energy recovery to reduce the unit cost of a treatment system. On 608 the other hand, COD removal strongly affects unit costs of the main unit process as 609 shown in Figures 22. The unit cost of UASB systems decrease with increasing COD 610 removal. The unit capital cost ($/reactor volume m 3 ) of UASB is much higher than that 611 of activated sludge tank for small-sized WWTPs, as shown in Table 9. The total cost 612 of treatment system is significantly by the cost of UASB. Although textile wastewater 613 has the highest energy recovery ratio, the COD concentration of textile wastewater is 614 lower than that of other wastewaters. This leads to higher unit costs of the UASB. The 615 meat processing wastewater system consumes more oxygen for removing N in PN/A. 616 However, this process increases the CH4 production of AD through increasing biomass. 617 Higher biodegradability and pollutant concentration reduce the unit cost of the meat 618 processing wastewater treatment system. Therefore, it is more economical to treat meat 619 processing wastewater with innovative technologies.

628
In the current study, an Excel-based model was developed to compare the unit 629 energy and cost of energy positive systems for treating different kinds of industrial 630 wastewater. Results showed that micro-sieving could remove 35.36%, 39.08%, and 631 32.50% of COD for tannery, meat processing, and textile wastewater, respectively. Due 632 to the relatively high TSS/COD ratio, primary treatment could achieve high COD 633 removal efficiency of textile wastewater with low TSS concentration as compared to tannery wastewater. UASB reactor could remove 15.8%-53.5%, 14.0%-49.0%, and 635 22.9%-51.0% of COD for three kinds of wastewater, respectively. UASB converts 636 more COD to CH4 in meat process wastewater with high BOD concentration because 637 of the high BOD/COD ratio. Pearson correlation analysis shows that the TSS/COD 638 ratio and BOD/COD ratio are the key parameters in determining COD removal and 639 methane conversion rate of primary treatment and secondary treatment. These 640 regression equations could be used to predict COD removal by the primary and 641 biological processes as follows: 642 COD removal (%) =75.69(BOD/COD ratio) +57.53(TSS/COD ratio) +23. 15 R 2 =0.929 For meat processing wastewater, tannery wastewater, and textile wastewater, the mean 643 unit energy consumptions were 1.49, 1.37, and 1.39 kWh/kg N removal, which are close 644 to actual unit energy consumption of PN/A process. Mean unit energy production for 645 three types of wastewater in innovative system are 1.80, 1.77, and 1.73 kWh/kg BOD 646 removal, respectively. The energy positive system for treating meat processing 647 wastewater could produce more energy for removing the same amount of BOD than 648 other than other industrial wastewater treatment systems. The average unit cost for 649 meat processing wastewater, tannery wastewater, and textile wastewater are 0.54, 0.57, 650 and 1.12 $/kg COD removal, respectively. Textile wastewater treatment has the highest 651 energy recovery ratio. However, the low COD concentration of textile wastewater 652 leads to high unit cost of system. Meat processing wastewater has higher COD 653 concentration and biodegradability than other wastewaters, which reduce unit cost of 654 UASB and increases electricity production through increasing COD loading rate to 655 UASB and AD. A low unit cost of UASB and a high CH4 production rate improves the 656 unit cost of meat processing wastewater treatment systems. Therefore, it is more economical and easier to treat meat processing wastewater with micro sieving, UASB, 658 and PN/A technologies than other industrial wastewaters. Due to high energy recovery, 659 innovative technologies could effectively reduce effect of industrial wastewater 660 treatment on the global warming. However, meat processing wastewater treatment with 661 innovative technologies shows high eutrophication potential because of its high non-662 biodegradable COD concentration. UASB with anammox treatment for treating meat 663 processing wastewater could not be considered as the sustainable option when the 664 eutrophication criterion and discharge requirement are prioritized. Therefore, physical-665 chemical treatment may be required in the future food processing wastewater treatment 666 design to remove non-biodegradable COD. The economic advantages and limitations 667 of innovative technologies in treating industrial wastewater have been quantified by 668 using unit energy and cost per kg pollutant removed which provide valuable references 669 for future sustainable design of industrial wastewater treatment systems. 670

Availability of data and materials 672
The data supporting the conclusions of this study are available within the article. 673