Full-scale semi-centralized wastewater treatment facilities for resource recovery: operation, problems and resolutions

Since the 19th century, the conventional centralized wastewater collection and treatment system has been widely used, and obtained a remarkable effect in minimizing pollutant discharge. However, the highly energy-intensive wastewater treatment facilities are often located in the downstream of urban rivers, and mainly aimed at reducing pollutant discharge. Consumption of limited resources is aggravated by difficult reuse of reclaimed water and low reuse efficiency (Libralato et al. 2012). Domestic wastewater can be divided into greywater (bathrooms, laundry and washing-machine wastewater) and blackwater (feces, urine, toilet paper and flush water), and whether kitchen wastewater belongs to greywater or blackwater depends on sources and degrees of pollution. Greywater and blackwater are collected and treated separately to achieve wastewater recycling and energy and nutrient (like carbon, nitrogen and phosphorus) recovery (Lam et al. 2015). Source separation of wastewater, quality-separation treatment and resource recovery are important trends of wastewater treatment. Nowadays, treatment and resource recovery technologies of greywater and blackwater are becoming mature (Chrispim & Nolasco 2016; Fumasoli et al. 2016; Wasielewski et al. 2016), there have been some application cases of wastewater source-separation and resource recovery in small communities and single buildings, and its practicability has been explored in larger residential areas in recent years (Hao et al. 2010). Semi-Centralized Resource Recovery Center (SCRRC) refers to the greywater and blackwater being collected and treated separately in a certain area, so as to meet the water quality requirements of different reuse goals, and its scale is between the traditional centralized and the decentralized systems (Chen et al. 2011; Tolksdorf et al. 2016). Biogas production through co-digestion of sludge and biowaste/urban organic matter recovers calorific energy in wastewater. Moreover, the scale and the number of modules can be flexibly adjusted to meet the requirements of the rapid development of urbanization. This paper takes an SCRRC project as a case study, and the feasibility of the full-scale application of semi-centralized wastewater treatment and the resource recovery system are investigated with productive experimental data.

The first full-scale SCRRC was implemented for the Qingdao Horticultural EXPO from 2014 (Tolksdorf et al. 2016). Figure S1 in the Supplementary Data is a photograph of the SCRRC. After planning and design, the full-scale demonstration project was constructed in a 2 km² region with a total of 700,000 m² building area. As shown in Figure 1, in the region, domestic wastewater is separated into blackwater and greywater at source, and then transported in separate pipes to two different treatment lines. Reclaimed water, electrical power and sludge fertilizer are recycled nearby. The catchment area consists of two communities, three-star hotels, a research institute, an enterprise headquarters and other public facilities. The wastewater source-separation pipelines were chosen at the planning and design phase, and domestic wastewater was divided into lightly polluted greywater from bathrooms, wash-basins, and washing machines, and blackwater, including kitchen and toilet wastewater. Greywater and blackwater are sent by corresponding drainage pipes to the treatment center to be purified, and disinfected reclaimed water is delivered to the user point through the reclaimed water pipes. In addition, kitchen waste from the surrounding canteens is collected and transported to the treatment center and is co-digested with sludge after pretreatment for biogas production. The three material streams are collected and transported to the treatment center, independently, and processed by greywater, blackwater and biological waste and sludge disposal modules, separately. The three modules are demonstrated below (Figure 1), respectively.

Greywater of each building is transported to the treatment center through gravity-flow pipeline. This project is designed to treat 700 m³ greywater per day. The membrane bioreactor (MBR) process is the core of the greywater units, and the processing of greywater units is shown in Figure S2. MBR is a promising technology for greywater treatment and reclamation because: (1) the effluent can satisfy the criteria set by USEPA for water reuse, (2) it has no plant-size limitation, and (3) the physical filtration intensifies the stability of the treatment systems (Hu et al. 2014).

The MBR is composed of a two-stage biochemical and two-cell membrane tank, and the design parameters are as follows: total effective volume = 275 m³, HRT = 9.4 h, influent loading rate = 0.027 kg BOD₅·(kg MLSS·d)⁻¹. Because of the low content of N and P in greywater, the effluent after subsequent disinfection can serve as flushing water for surrounding residential areas, office areas and hotels, and the extra reclaimed water can be used for road cleaning, car washing (GB/T 18920-2002) or supplying to landscape water bodies (GB/T 18921-2002).

Considering the possibility of misconnection among greywater and blackwater pipelines in some site points, two operation states are set up in the greywater biochemical tanks: stirring (anoxic and anaerobic) or aerating (aerobic). If influent chemical oxygen demand (COD), N and P are quite low, complete aerobic treatment is adopted in the biochemical tanks for COD and ammonium oxidation. If misconnection occurs, with toilet wastewater and kitchen wastewater flowing to greywater, the biochemical tanks convert to the anoxic/anaerobic state for denitrifying and orthophosphate release, and the aerobic state in the membrane tank can provide denitrifying and orthophosphate uptake. The liquor return and waste flowrate from the membrane tank is controlled for strengthening biological nutrient removal.

Blackwater has a high concentration of organics and nutrients. Anaerobic treatment with biogas production has been used for blackwater treatment and resource recovery (Kujawa-Roeleveld et al. 2006; De Graaff et al. 2010). Blackwater of each building is transported to the treatment center through gravity-flow pipelines. The design treatment capacity is 800 m³/d, and the AAO-MBR (anoxic/anaerobic/aerobic bioreactor with liquor and solids separation by membrane) process is the core process of blackwater treatment. The AAO-MBR unit is divided into two groups and can be operated in parallel. The total effective volume of each biochemical tank is 628 m³, and the designed HRT is 37.7 h. The effluent after deep treatment (activated carbon adsorption and chlorination) is mainly used as gardening and irrigation water, landscape replenishment and road-washing water in the surrounding area according to the standards (GB/T 18920-2002; GB/T 18921-2002). The blackwater treatment process is showed in Figure S3.

The blackwater treatment module receives high-strength reject water (the digested sludge pressure filtrate) from the anaerobic sludge treatment module. In order to improve the nitrogen removal efficiency, the biochemical tank is divided into six stages: the first, second, and sixth stages are in the anoxic/anaerobic state; the fourth and fifth stages are in the aerobic state; the third stage can switch between anoxic and aerobic states. A ratio of mixed liquor from the membrane tank is returned to the fourth stage firstly for consuming the high dissolved oxygen concentration that results from strong aeration in the membrane tank, and then the returning liquor for denitrifying is transported from the end of the fifth stage to the first stage. Moreover, the returning ratios are adjustable. External carbon source (sodium acetate) addition and chemical phosphorus removal are needed for satisfying the discharge standard Grade 1A (GB/T 18918-2002). In addition, due to the high chromaticity of blackwater and reject water (Abegglen et al. 2009), an activated carbon filter tank is installed as an optional process for controlling the chromaticity of the reclaimed water.

Organic matter in kitchen waste accounts for over 95% of the dry matter, which has great potential in energy recovery (Chen et al. 2013; Zhang et al. 2017; Zheng et al. 2018). In this module, after appropriate pretreatment, the collected kitchen waste is anaerobically co-digested with primary sludge and excess sludge from the blackwater module and excess sludge from the greywater module for biogas production. The biogas is used for co-generation to recover power and heat, which can realize the highly efficient utilization of the potential energy in the waste.

The anaerobic digestion tank is comprised of two separated olive-shape anaerobic chambers. The effective volume of the single tank is 553 m³, the design slurry volume is 51.6 m³ per day (30.0 m³ of sludge and 21.6 m³ of kitchen waste), and the design anaerobic digestion time is 20 d. The total daily gas production under full load is expected to reach 2,100 m³·d⁻¹.

The digested sludge is periodically discharged to the digested sludge tank, and then dewatered by a recessed plate filter press for disposal. The pressure filtrate can be treated independently by the side-stream sequencing batch reactor (SBR) and returned to the storage tank of the blackwater treatment module.

Sampling information and analytical methods for physical and chemical parameters are detailed in the Supplementary Data.

As shown in Table S1 (Supplementary Data), the actual influent COD, TN and TP exceeded the design values of the greywater system. The reason for this phenomenon is that a part of the blackwater and greywater were misconnected during the construction period and this means a content of high-concentration blackwater flows into the greywater module (Ren et al. 2018; Tolksdorf et al. 2018). According to the utilization routes and required water quality standards (GB/T 18920-2002) of the purified greywater, the biochemical tank is kept in an anoxic/anaerobic state, creating an anaerobic/anoxic/aerobic environment with aeration in the membrane tank. Furthermore, the mixed liquor in the membrane tank is returned to the inlet of the biochemical tank. The measured dissolved oxygen (DO) was <0.2 mg·L⁻¹ in the biochemical tank, and 7–8 mg·L⁻¹ in the membrane tank. Hence, biological denitrification and phosphorus removal were achieved in the greywater treatment module.

Operation results (Figure S4, Supplementary Data) shows that influent COD of the greywater ranged from 350 to 800 mg·L⁻¹, while effluent COD was lower than 40 mg·L⁻¹, and notably, under >94% probability, effluent COD was lower than 30 mg·L⁻¹. The average influent COD was 496 mg·L⁻¹, while the average effluent COD was 17.4 mg·L⁻¹, and the COD removal efficiency was about 94.9% (Figure 2). It is found that the actual influent COD concentration was higher than the design value (by up to about four times) due to the existence of misconnection in the greywater pipeline, whereas the daily flowrate of greywater was only 7.14–21.4% of the design value, and the actual influent organic loading rate was only about 34% of that, so that the effluent COD concentration could still reach a low level (lower than the Grade 1A standard value in GB/T 18918-2002).

As shown in Figure 3(a), the greywater module exhibited a good NH₃-N removal effect: average influent NH₃-N was 41.1 mg·L⁻¹, while the effluent NH₃-N was 1.36 mg·L⁻¹, and the average removal efficiency was about 96.5%. Overall, the greywater modules showed good biological nitrification ability in most conditions, although some outliers existed (Figure S5(a)). It is shown according to the operation results that, under 7% probability, effluent NH₃-N exceeded 5.0 mg·L⁻¹. The ammonium loading rate was 1.8 times higher than that of design and fluctuated greatly (Figure S5(a)), but the biological nitrification could be reacted thoroughly owing to high sludge concentration and plenty of dissolved oxygen in the membrane tank.

Figure 3(b) demonstrates that the average influent TN of the greywater module was 54.2 mg·L⁻¹, while effluent TN was 8.48 mg·L⁻¹, and the removal efficiency was about 81.4%. Meanwhile, influent TN loading was 6.8 ± 4.4 kg/d, which was 1.94 times higher than the design value. The biological nitrification and denitrification exhibited good efficiency in most conditions (Figure S5(b)) after adjusting the aeration zone to the anoxic agitation mode in the biochemical tank, even with the existence of misconnection among blackwater and greywater pipelines.

As shown in Figure S6(a), as a result of few chromogenic pollutants in the greywater, most could be removed by biochemical treatment, and the chromaticity of the effluent was in the region of 15–20 PCU, which satisfied the chromaticity requirement (≤30 PCU in GB/T 18920-2002) for the flushing water of toilets. For the MBR membrane units, we chose a flat membrane module with a pore size of 0.04 μm. The high-efficiency retention of the membrane led to the low content of effluent suspension, and the effluent turbidity was less than 0.7 NTU, which satisfied the water quality requirements (GB/T 18920-2002) of recycled water for flushing.

The influent flowrate of the blackwater module generally was 200–350 m³·d⁻¹, and only one group of biochemical and membrane tanks was then operating as a result of insufficient influent. The influent COD concentration of the blackwater was slighter lower than that of design due to the misconnection. In general, the influent pollutant concentration of the blackwater module was lower than the design value owing to the misconnection between the blackwater and greywater pipelines. Meanwhile, the liquid volume was unable to reach a high level. The influent COD loading of the blackwater was 230 ± 90 kg·d⁻¹, which was 44.6% of that of a single production line under the design conditions. The blackwater module exhibited good COD removal performance. It is shown in Figure 4 that the average influent and effluent COD were 813 mg·L⁻¹ and 21.6 mg·L⁻¹, respectively, and the average removal efficiency could reach 97.2%.

Figure S7 illustrates that effluent COD exceeded 40 mg·L⁻¹ not more than 11 times during the operation monitoring period. Furthermore, with under 99.1% probability, the effluent COD could be kept below 40 mg·L⁻¹ and satisfied the discharge standard Grade 1A (GB/T 18918-2002).

Figure 5 and Figure S8 reflect the NH₃-N and TN removal efficiency of the blackwater module with dosing of external carbon source. Influent TN concentration was relatively stable from 2016 to 2017, the average influent and effluent TN were 108 mg·L⁻¹ and 23.0 mg·L⁻¹, respectively, and average removal efficiency was 77.8%, whereas the influent TN had obvious seasonal variation from 2017 to 2018 (Figure S8(b)), a large amount of wastewater and low concentration of pollution in summer and autumn, and the lowest pollution value was 51 mg·L⁻¹, while it was the opposite in spring and winter, and the highest value was 160 mg·L⁻¹. The average influent and effluent TN were 115 mg·L⁻¹ and 23.8 mg·L⁻¹, respectively, and average removal efficiency was 78.8%. In addition, because the blackwater flowrate fluctuated greatly and the operation of the blackwater module was unstable from November to December 2017, the effluent TN of the MBR was high (20–40 mg·L⁻¹), and the removal efficiency was only 70–80% (Figure S8(b)). By optimizing the dosing point of sodium acetate in 2018, i.e., the dosing position was changed from the inlet end to the post-anaerobic zone (the sixth stage) of the biochemical tank, the utilization efficiency of the external carbon source was improved (Cui & Liu 2015). Thus the effluent TN was reduced to within 15 mg·L⁻¹ and the removal efficiency was over 92% (Figure S8(b)).

The trend of NH₃-N concentration change was similar to that of TN in the blackwater module. The NH₃-N was effectively removed by the blackwater module. Average influent and effluent NH₃-N of the blackwater were 93.9 mg·L⁻¹ and 2.86 mg·L⁻¹, and the total average removal efficiency was 97.3% (Figure 5(a)).

The effluent turbidity of the blackwater module was generally lower than 0.3 NTU, and the maximum was no more than 0.6 NTU. The average effluent chromaticity was 19.6 PCU, and only 8% of data (Figure S6(b)) was higher than the limit value (30 PCU) in GB/T 18920-2002. The main reason for this is that the blackwater had plenty of organic pollutants and chromogenic groups. Moreover, the chromaticity of the dewatered solution after anaerobic digestion was high, and was hard to completely remove, even through biochemical treatment process, so that the effluent still contained a small amount of chromogenic pollutants, and there was high chromaticity in the effluent accordingly. At present, the chromaticity of the reclaimed water (purified blackwater) can meet the water requirements of road sprinkling and greening (GB/T 18920-2002), so it is mainly used to supply for gardening and road sprinkling in surrounding areas and there is no need to arrange the activated carbon filter tank of the blackwater module.

From the anaerobic co-digestion and biogas production units being started up, it was hard to guarantee the amounts of kitchen waste, only the primary and excess sludge generated from the treatment of greywater or blackwater used as anaerobic digestion substrates, and biogas was converted into heat energy by the boiler to maintain the digester temperature. High-temperature anaerobic digestion (55 °C) operation mode was adopted from October 2017 until mid-December and gradually transformed to the mesophilic anaerobic digestion mode, and at last the digestion temperature stabilized at the range of 34–36 °C in February 2018. The sludge feed amount of the anaerobic digestion tank was only 5.4–5.6 m³·d⁻¹ (water content 97–98%), so only one anaerobic digestion tank was operated.

As shown in Figure 6, the daily biogas production could reach 40–60 m³·d⁻¹, and the average gas production rates could reach 10.3 m³ gas·(m³ raw sludge)⁻¹ with a methane percentage of over 70% under high-temperature operating conditions. If the feed increased, the gas production raised correspondingly, indicating that the operating loading of the anaerobic digestion tank was low under the current feeding condition, and the total amount of organic matter entering the anaerobic digestion tank was the main factor affecting the gas production.

In order to investigate biogas production by the anaerobic co-digestion of sludge and urban organic matter, 100 L·d⁻¹ waste molasses was added to the raw sludge tank (amounting to COD concentration 635 g·L⁻¹) from March 2018 under the condition that the sludge feed and digestion temperature were constant, and maintained for two months, to explore the variation in gas production efficiency. As shown in Figure 6, the biogas production rose from 50–60 m³·d⁻¹ to 100–140 m³·d⁻¹ and stabilized after the adding of molasses, at about twice as much as that without adding molasses. Furthermore, the addition of molasses had no significant effects on the methane content. In terms of material analysis, it was found that the COD total amount of 100 L molasses was similar to that consumed by 5.5 m³ raw sludge under normal operating conditions (1:1.16), so the gas production per unit organic matter was doubled. Furthermore, after the addition of waste molasses, the volatile solids (VS) degradation efficiency of the anaerobic digestion tank increased by 10%, and there was no distinct change in the VS content of the digested sludge, revealing that the organic matter of the waste molasses could be fully utilized. As a by-product of sugar production, molasses was used as a substrate of the anaerobic co-digestion with the sludge, while the sludge properties would not change after co-digestion, so that molasses as an auxiliary material in anaerobic digestion could realize the resource and energy recovery of the waste.

The main index of reclaimed greywater could meet multiple water demands, and it has been used for landscape, gardening and miscellaneous uses in surrounding areas in spring, summer and autumn. Since the blackwater module received reject water (pressure filtrate from sludge dewatering) which had high concentrations of unbiodegradable soluble organic matter, the chromaticity of the reclaimed blackwater was high and was mainly used for gardening and road cleaning in surrounding areas. Statistics indicated that the consumption of recycled water in winter was small, while the utilization rate of recycled water reached over 50% from March 2018 (Figure S9).

After conditioning with polyacrylamide (PAM) and ferric chloride, digested sludge was dewatered by recessed-plate filter press. Monthly characteristic data of sludge cake (water content 78–80%) are shown in Table S2.

As displayed in Table S2, in terms of the dewatered anaerobic digested sludge, the average content of organic matter reached 555 g·kg⁻¹ sludge, and total content of N, P and K reached 120 g·kg⁻¹ sludge, which all meet Grade A and Grade B of the control standard for agricultural use (CJ/T 309-2009). In addition to the total mercury, heavy metals, various microbial, nutritional, health and other indexes could meet the agricultural sludge standards A level (CJ/T 309-2009). It can be seen that dewatered sludge cake could be applied directly to green belts as well as economic crops other than grain and vegetables. Sludge cake produced by the treatment center was taken away by surrounding farmers for soil and forest improvement. The dewatered sludge could be used for agriculture through proper composting, and in this way could not only adjust the soil properties and realize waste utilization, but also reduce fertilizer consumption (Abbasi et al. 2019; Brisolara & Bourgeois 2019).

The operating costs of SCRRC were comprised of power and chemicals. As shown in Figure S10, the cost of power consumption occupied a majority share of the total costs per ton of treated water, with a proportion of 55.3–94.2%, and the cost of chemicals was only 5.8–44.7%.

Power user units could be classified to three modules: the greywater module, the blackwater module, and the sludge and biogas module. The electrical devices are shown in Table S3, and the detailed distributions of power consumed by the different modules are presented in Figure S11. Figure S11 shows that the blackwater module used the most power followed by the sludge and biogas module, while the greywater consumed the least. The proportion of power usage was ranked as 53.7% > 32.1% > 14.3%. Because of low pollutant concentrations in the influent, the greywater treatment line was short with fewer electrical devices, and an odor removal system was not installed. The electrical equipment was divided into three parts, influent pre-treatment, bioreactor system and membrane system, with power consumption share of 15.6%, 12.9% and 71.4%, respectively. To reduce membrane fouling for the good operation of the membrane modules, over-air was supplied in the membrane tank, which resulted in the highest power consumption. Compared with the greywater module, the blackwater module had a longer treatment line with more electrical equipment. Because the full plant is a closed-room type, odor should be removed from the treatment process of the blackwater. Hence, power consumption in the blackwater module was classified into influent pre-treatment, bioreactor, membrane system, chemical dosing system and odor removal system, and the proportions were 7.7%, 25.5%, 35.5%, 2.7% and 28.7%, respectively. The cost of aeration for bioreactor and membrane tank was as high as ∼60%. As for the sludge and biogas module, the mass flow is more complicated than wastewater treatment, so the electricity devices were grouped into more parts: sludge cycling, dewatering, sludge cake transportation, leachate system, biogas system, and odor removal, with power consumptions of 27.2%, 1.1%, 0.7%, 3.7%, 19.4% and 48.0%. It should be noted that the odor removal process occupied almost half the total power cost in the sludge and biogas module.

Wastewater source-separation should be implemented in all buildings for this project from the planning and construction stage, specifically the greywater and blackwater should be collected separately, and the inspection well of greywater and blackwater should be set out of doors, respectively, and then transported to the treatment center by independent pipelines. After the project was completed and put into operation, from the measured water quality of greywater and blackwater, it was found that misconnection occurred between blackwater and greywater pipelines, namely, that the concentration of greywater had increased and that of blackwater was lower than the design value. By sampling and analyzing the blackwater and greywater from residential quarters, hotels and public buildings, it is worth noting that the overall misconnection rate was 27.9% (Ren et al. 2018). Further investigation revealed that a five-star hotel in the project service area partially realized source-separation, but mixed outdoors, in that the domestic wastewater in the area all entered the blackwater pipeline. Furthermore, a reverse connection occurred when the greywater and blackwater pipelines of a commercial office building merged into the individual main pipes of the road. In addition, most buildings had not been furnished in a newly built residential area, and the owners connected the drainage pipelines indiscriminately during decoration, and the phenomenon of misconnection in this community was prominent. In contrast, misconnection was not found in the pipelines of two three-star hotels, finely decorated residential areas and a commercial building, so that typical water quality characteristics of greywater and blackwater were presented. In terms of the actual operation effects of the project, although the main pollutant concentration in the influent of the greywater unit was higher than the design values, the water quality index of the effluent was very low except for the increase of chromaticity, and there was almost no difference from that of purified pure greywater effluent. After disinfection, the effluent could be used as high-quality recycled water to meet sensitive water needs such as toilet flushing water in residential areas and ecological replenishment (Tolksdorf et al. 2016).

In light of the practice of the project, it is necessary to carry out more systematic work on the aspects of planning and design, construction, technical specifications and standards to achieve complete wastewater source-separation, independent collection and treatment and resource utilization in a certain region.

High TN concentration occurred in the effluent of blackwater units occasionally, especially when the pressure filtrate entered the treatment unit. The blackwater of this project came from toilet-flushing water and kitchen wastewater, and the time of blackwater from production to transportation to the treatment center was only a few hours. Analyzing COD components of the blackwater regulation tank (Table S4), it is found that the particulate COD (pCOD) accounts for about 51.6% of total COD (tCOD), whereas the proportion of colloidal COD (cCOD) and soluble COD (sCOD) was only 48.4%. After conditioning and primary precipitating, the tCOD of the blackwater decreased to 400–600 mg·L⁻¹, the average tCOD/TN was 4.09–5.45, and the sCOD was about 250 mg·L⁻¹. The carbon source for the biological denitrification was lower, which has an adverse impact on biological denitrification. In addition, the pressure filtrate entering the blackwater system increased the nitrogen loading of the influent. Approximately 130 mg·L⁻¹ external carbon source was added in the blackwater module so that it could satisfy the discharge standard Grade 1A (GB/T 18918-2002), accordingly.

Based on the blackwater characteristics shown in Table S4, we proposed a combined process to recover the carbon source and provide good nitrogen removal efficiency: carbon capture, side- fermentation and acidification, and mainstream short-cut biological nitrogen removal. We carried out an investigation on carbon recovery from blackwater. We used a chemical enhanced high-rate activated sludge process (CEHRAS) (Jiang et al. 2019) to capture and concentrate the carbon source, and we did a feasibility assessment of whether volatile fatty acids (VFAs), produced from anaerobic fermenter feed with captured carbon, could be used for short-cut nitrogen removal. It was found that the best carbon capture efficiency of CEHRAS was 75% with 30 mg·L⁻¹ Fe salt dosage, and the mass balance of COD and TN is shown in Figure S12.

The results show that while the nitrogen removal proportion over nitrite was 87.7%, a VFA production of the anaerobic fermenter of 506 g COD_VFA per kilogram sludge feed would meet the requirement for elector donors to remove nitrogen to 15 mg/L in the effluent.

Due to lack of kitchen waste, primary and excess sludge were the main substrate in the anaerobic co-digestion system for this project. Accordingly, the insufficient biogas was only used to heat and maintain the temperature of the digestion tank, while the remaining biogas was not enough to operate the cogeneration unit. During the period of anaerobic co-digestion of kitchen waste and sludge, it was found that the sorting of toothpicks, paper towels, plastic film and other ingredients was not thorough, and these easily blocked the pipeline and increased the difficulty of operation and maintenance. At the same time, shells and other components in the food residues not only affected the efficiency of the pretreatment equipment, but also caused wear in the equipment and piping after entering the anaerobic digestion tank. Therefore, the principle of classification should be implemented in the collection process of kitchen waste, and it is necessary to improve the removal efficiency of these components in the pretreatment units to avoid them entering the subsequent anaerobic digestion facilities.

In addition to kitchen waste, other municipal organic matter can also be the substrate of anaerobic co-digestion for biogas production with sludge after enhanced pretreatment, such as vegetable tails, food fermentation waste and grass organic matter.

Modular processing units are adopted in the treatment center, so how to realize cooperativity and compatibility among the processing units remains to be deeply explored. At present, the greywater and blackwater units could serve as safeguard methods for each other when in emergency conditions. Nevertheless, the anaerobic digestion system was hard to be guaranteed, for example, if the SBR failed, the pressure filtrate with high NH₃-N would enter the blackwater module directly, which would cause serious loading shock and lead to the effluent TN exceeding the standard, so that the process parameters would need to be adjusted in time.

The comprehensive operating cost of the project is higher than that of a conventional wastewater treatment plant. This project has lots of processing units and equipment for wastewater and biosolid treatment and resource recycling. Especially, the high energy consumption of the odor collection and treatment system and blowers for aeration made a great contribution to the cost. On the other hand, the dosing of the external carbon source, for blackwater and reject water treatment to meet the total nitrogen discharging standard, increased the total cost. However, the cost is acceptable considering the comprehensive costs of wastewater treatment, advanced wastewater treatment, and sludge treatment and stabilization, as well as ecological and social benefits.

In light of the wastewater and solid waste treatment, sludge anaerobic digestion and biogas utilization module involved in the project, if the scale is too small, the energy consumption and other costs of the system will increase. However, the energy consumption and chemical costs per unit of treatment capacity or sludge volume will also decline accordingly with the increasing treatment scale. According to previous research results of our group (Chen et al. 2011), the optimal scale of the SCRRC was 50,000–100,000 PE, which was equal to 10,000–20,000 m³·d⁻¹ of the overall treatment scale of greywater and blackwater. In this case, the corresponding investment and operation costs of the water supply and drainage pipe network and treatment facilities are all in the optimal area.

Semi-Centralized Resource Recovery Center can not only treat and reuse the wastewater nearby, and realize the recycling of water resources, but also achieve anaerobic co-digestion of sludge and waste or urban organic matter, so as to obtain biogas for self-energy replenishment, and thus actualize the recovery and high-efficiency utilization of the resources in wastewater to a certain extent. In order to realize the efficient reclaiming and reusing of resources and energy, and make carbon neutralization a reality, it is necessary to carry out further integration and application research in the further application of the semi-centralized resource recovery system based on source separation.

We gratefully acknowledge the financial support of this work from National Key Research & Development Plan of China (2017YFC0403402), and Key Intergovernmental (Sino-German) Scientific and Technological Innovation Cooperation Projects (2016YFE0123500).

The Supplementary Material for this paper is available online at https://dx.doi.org/10.2166/wst.2020.169.