The DHS concept is deceptively simple. Instead of forcing air through wastewater in energy-hungry aeration tanks, the system lets wastewater trickle downward over stacked cubes of polyurethane sponge, each enclosed in a cylindrical net frame. The sponges act as biomass carriers, hosting dense microbial communities that degrade organic matter and convert ammonium to nitrate. Because the biomass clings to the sponge rather than floating freely, the reactor retains very high sludge concentrations with sludge retention times exceeding 100 days, and the oxygen the microbes need is supplied passively by natural ventilation rather than by blowers. Oxygen gradients develop within each sponge cube, creating an aerobic outer layer and oxygen-limited inner zones where denitrification can proceed — meaning a single reactor can perform both nitrification and partial nitrogen removal.

What makes the Susaki installation scientifically significant is its configuration and its context. The plant replaced an extended-aeration activated sludge process originally designed in 1976 for 1,800 cubic meters per day. With Susaki’s population shrinking and little prospect of expanding sewer connections, the new DHS system was deliberately designed for just 500 cubic meters per day, with an average operational flow of 365 cubic meters per day. The reactor comprises ten identical units operated in parallel, each consisting of four stacked segments filled with sponge carriers, giving a total sponge volume of 64 cubic meters and a theoretical hydraulic retention time of roughly 4.2 hours. Downstream, an aerated moving bed biological filter filled with anthracite media polishes the DHS effluent before final chlorination and discharge into the Mitarai River.

The monitoring program was rigorous. From June 2021 through December 2023, researchers from the National Institute for Environmental Studies and their collaborators collected weekly grab samples of raw sewage and DHS effluent, tracking dichromate-based chemical oxygen demand, biochemical oxygen demand, total suspended solids, ammonium, nitrate, total nitrogen, and fecal indicator bacteria. The influent was typical medium-strength domestic sewage, with a chemical oxygen demand of 368 milligrams per liter and a biochemical oxygen demand of 134 milligrams per liter. Against this feed, the DHS reactor consistently delivered effluent with 77 to 85 percent removal of organic pollutants and a 1.5 log10 reduction in Escherichia coli, comfortably meeting Japanese discharge standards even before the polishing stage. The full treatment train pushed final effluent to a biochemical oxygen demand of just 9 milligrams per liter and ammonium of 3 milligrams of nitrogen per liter.

Winter posed the first stress test. Between December 2021 and March 2022, wastewater temperatures generally stayed below 20 degrees Celsius and influent chemical oxygen demand rose. Organic removal dipped slightly — effluent chemical oxygen demand crept from 45 to 66 milligrams per liter — but the downstream filter kept the plant compliant, and ammonium oxidation remained comparatively stable. This resilience matters because most previous full-scale DHS experience had been gathered in tropical or subtropical conditions, in countries such as India, Egypt, and Thailand, typically as a post-treatment step after anaerobic reactors. Susaki provided the first extended evidence that the technology can weather a temperate winter while directly treating raw domestic sewage.

The second test was deliberate self-shrinking. From April to December 2022, the operators intermittently shut down one of the ten units in weekly rotations, reducing the active sponge volume to 57.6 cubic meters and the theoretical retention time to 3.8 hours, simulating the capacity reductions that declining inflows will eventually demand. Effluent quality barely budged: chemical oxygen demand, biochemical oxygen demand, and ammonium remained essentially indistinguishable from full-capacity operation. Laboratory experiments on sponge carriers removed from service showed that roughly 30 percent of the attached sludge was lost over eight days of feed stoppage, driven by endogenous respiration — microbes digesting their own reserves — yet the volatile fraction of the sludge recovered within two weeks of restarting, indicating rapid biomass regrowth. The parallel-unit architecture thus functions as a dial, allowing operators to match treatment capacity to demand without sacrificing performance.

The third modification was almost embarrassingly low-tech. In March 2023, engineers slid a polyurethane sponge sheet about 20 millimeters thick directly beneath the wastewater distributor at the top of the reactor, above the first sponge-filled segment. The intent was to spread the incoming flow more evenly and eliminate channeling. The effect on nitrification was dramatic. Ammonium in the DHS effluent fell from 16 to 19 milligrams of nitrogen per liter to about 9, nitrate accumulated in the lower segments as expected from intensified oxidation, and statistical analysis confirmed that ammonium removal efficiency rose significantly — from 37 percent during normal operation to 66 percent — even though the ammonium loading itself did not differ among periods. Dissolved oxygen across the reactor segments climbed from 4.2 to 5.0 milligrams per liter to 5.5 to 7.1, and oxidation-reduction potential shifted from slightly negative to strongly positive values. Tracer tests showed the measured hydraulic retention time rising from 89 to 92 percent of the design value, confirming improved flow dispersion.

The energy accounting is where the story becomes genuinely striking. Under the old activated sludge process, aeration blowers, foam water pumps, and diffused-air equipment dominated electricity use, and because a minimum aeration rate must be maintained regardless of flow, the specific power consumption rose as flows fell. In 2016 the conventional process consumed 1.803 kilowatt-hours per cubic meter of treated water. The DHS system, by contrast, used 0.473 kilowatt-hours per cubic meter in 2023, and excluding equipment shared by both configurations, the process-specific figure was 0.270 kilowatt-hours per cubic meter — approximately 83 percent lower. Sludge handling improved too: dewatered sludge cake production fell by half, from 6,195 to 3,145 kilograms per month, aided by better dewaterability and cake moisture dropping from 82 to 74 percent. Macrofauna such as drain fly larvae colonizing the sponge media likely contributed to sludge reduction through predation, a quirk of trickling-filter ecology that the researchers note deserves further study.

The broader significance extends well beyond one small Japanese city. Japan’s sewerage network now covers 93.7 percent of the population, but roughly 30 percent of the country’s 5,353 rural community treatment facilities have operated for more than three decades, and nationwide sewage volumes have been falling since fiscal year 2015 as regional populations decline. Conventional activated sludge and membrane bioreactor systems fix their capacity in tank volume, making it difficult to cut costs and energy use when inflow drops. The DHS system offers a different model: modular units that can be switched off in rotation, passive aeration that slashes electricity demand, biomass that survives outages and recovers quickly, and a reactor height of about five meters that conceptually fits within the footprint of existing aeration tanks during facility renewal. Maintenance requirements proved modest — mainly annual manual cleaning of distributor ports to prevent clogging.

The authors are careful about limits. The findings come from a single full-scale plant under site-specific conditions, microbial communities were not analyzed in depth, and the previous facility’s oversized equipment somewhat flatters the energy comparison. Still, the thirty-month record from Susaki demonstrates something rare in wastewater engineering: a technology that gets smaller, cheaper, and cleaner in step with the community it serves. For thousands of aging facilities in Japan and countless decentralized plants worldwide facing shrinking budgets and shrinking populations, the humble sponge may prove to be an unexpectedly powerful answer.

Subject of Research: Long-term full-scale performance of a parallel-unit downflow hanging sponge wastewater treatment system in Japan

Article Title: Long-term monitoring of a full-scale parallel-unit downflow hanging sponge system with adaptive capacity for fluctuating wastewater flows in Japan

Article References: Takemura, Y., Yamazaki, S., Matsueda, T., Ohmori, S., Nagano, A., Taniwaki, M., Tomioka, N., Aoki, M., & Syutsubo, K. (2026). Long-term monitoring of a full-scale parallel-unit downflow hanging sponge system with adaptive capacity for fluctuating wastewater flows in Japan. Environmental Science and Pollution Research. https://doi.org/10.1007/s11356-026-38300-y

Image Credits: AI Generated

DOI: 10.1007/s11356-026-38300-y

Keywords: downflow hanging sponge, wastewater treatment, decentralized sewage treatment, energy efficiency, nitrification, excess sludge reduction, population decline, Japan, activated sludge alternative, hydraulic retention time, polyurethane sponge media, moving bed biological filter

Researchers Bernardo João Francisco Companhia and Paulo Sérgio Scalize of the Federal University of Goiás in Brazil set out to solve a problem that has dogged electrocoagulation, an emerging water treatment technology, for years: how do you take results from a small laboratory batch reactor and faithfully translate them into a full-scale, continuously flowing treatment plant? Electrocoagulation works by passing an electric current through water between sacrificial electrodes, typically made of aluminum. The anode dissolves electrochemically, releasing aluminum ions that hydrolyze in solution to form aluminum hydroxide flocs. These gelatinous, highly adsorptive particles sweep fluoride ions out of the water through adsorption and coprecipitation, while gas bubbles generated at the cathode float much of the sludge to the surface. The process needs no added coagulant chemicals, which is a major sustainability advantage over conventional chemical treatment, but its performance depends on a delicate interplay of current density, contact time, electrode geometry, and, crucially, the way water actually moves through the reactor.

The team built two cylindrical reactors with concentric aluminum electrodes, one operating in batch mode and one in continuous flow, and challenged both with synthetic water containing 10 milligrams of fluoride per liter, nearly seven times the WHO limit. In the batch phase, they tested three current densities: 9.79, 13.06, and 16.32 amperes per square meter. All three eventually drove fluoride below the guideline value within ten minutes, but the energy economics told a more nuanced story. The lowest current density achieved 95.6 percent removal for just 3.27 kilowatt-hours per cubic meter of water treated. Raising the current to 13.06 amperes per square meter lifted removal to 98.25 percent but nearly doubled the energy bill to 5.97 kilowatt-hours per cubic meter. Pushing further to 16.32 amperes per square meter added a mere 0.45 percentage points of removal while consuming 9.04 kilowatt-hours per cubic meter. The sweet spot, the researchers concluded, was the middle condition, which balanced near-complete defluoridation against a tolerable energy demand.

With the optimal batch parameters in hand, the team transposed them to the continuous flow reactor and varied the hydraulic retention time, the average time water spends inside the reactor, from three to seven minutes. The results were dramatic. At retention times of six and seven minutes, the continuous system removed more than 98 percent of fluoride, stabilizing residual concentrations at just 0.15 and 0.12 milligrams per liter respectively, far below the WHO guideline. Shorter retention times performed progressively worse: five minutes yielded about 81 percent removal, four minutes 60 percent, and three minutes only around 45 percent. But here is the headline finding: at every single retention time tested, the continuous reactor outperformed the batch system that had been used to calibrate it. The improvement ranged from 9.7 percentage points at seven minutes to a remarkable 17.7 points at three minutes, and paired statistical tests confirmed the differences were highly significant, with p values below 0.001.

The energy story was equally compelling. Specific energy consumption in continuous flow was lower than in batch mode across all conditions, with savings reaching up to 40 percent. At the six-minute retention time, the continuous system consumed roughly 1.40 to 1.45 kilowatt-hours per cubic meter, a fraction of the 5.97 kilowatt-hours per cubic meter recorded in the optimized batch run. The researchers attribute this advantage to the hydrodynamic regime of the continuous reactor. Under near plug flow conditions, water moves through the vessel in an orderly fashion, and freshly generated aluminum hydroxide coagulant is distributed uniformly along the reactor’s length. Each parcel of water encounters newly formed flocs as it travels, maximizing the contact between adsorbent and contaminant. In a batch vessel, by contrast, mixing is temporal rather than spatial, and the effective use of electrochemically generated coagulant may be limited by the way flocs and fluoride interact within a single, well-stirred volume.

To understand exactly how water behaved inside their continuous reactor, the team performed a residence time distribution analysis, a classic chemical engineering technique in which a pulse of tracer, in this case methylene blue, is injected at the inlet and its concentration tracked at the outlet over time. The tracer curve peaked between two and three minutes, with a long asymmetric tail stretching to roughly fourteen minutes, a signature of moderate axial dispersion. Quantitatively, the axial dispersion number came out at 0.154, indicating a predominantly plug flow regime with limited longitudinal mixing. The real hydraulic retention time measured by the tracer test closely matched the theoretical value calculated from reactor volume and flow rate, meaning the reactor delivered the contact time its designers intended, a precondition for any trustworthy scale-up. The slight deviation from ideal plug flow, the authors suggest, may actually help, enhancing contact between coagulants and contaminants without compromising overall performance.

Perhaps the most practically valuable contribution of the study is a quantitative bridge between the two operating modes. The researchers defined conversion factors, ratios relating removal efficiency and energy consumption in batch mode to those in continuous mode, and found that both factors decrease steadily as hydraulic retention time increases. The efficiency conversion factor fell from about 1.64 at three minutes to 1.11 at seven minutes, while the energy conversion factor dropped from 1.7 to 1.3. Both relationships fit second-order polynomial models with coefficients of determination above 0.98, meaning the convergence is strong and predictable. In plain terms, the longer the water stays in the reactor, the more the performance gap between batch and continuous operation closes, and the more reliably batch-derived laboratory data can predict full-scale continuous behavior. This gives engineers a mathematical tool for designing continuous electrocoagulation plants directly from bench-scale experiments, rather than relying on trial and error.

The chemistry underlying the process proved robust throughout. In batch experiments, the pH of the initially acidic water, which started at 4.32, spiked toward 8.2 in the first five minutes as hydroxide ions accumulated at the cathode, then settled into a stable range between 5.5 and 6.2, conditions known to favor fluoride adsorption onto aluminum hydroxide. In continuous operation, pH barely moved at all, hovering between roughly 5.3 and 6.8 regardless of retention time, a buffering effect arising from the balance between hydroxide generation at the cathode and hydroxide consumption during aluminum hydrolysis. Electrical conductivity declined modestly as ions were scavenged by flocs but otherwise remained stable, indicating that the treatment removes fluoride without substantially altering the water’s overall ionic character. The researchers note that they deliberately avoided adding supporting electrolytes or adjusting pH, testing the process under restrictive, realistic conditions rather than chemically favorable ones.

The implications extend well beyond fluoride. Electrocoagulation has demonstrated effectiveness against phosphorus, nitrates, organic matter, turbidity, dyes, and even resistant microorganisms, and the framework developed here, coupling electrochemical optimization with rigorous hydrodynamic characterization, offers a template for scaling any of these applications. The authors are careful about the limits of their work: sludge production was observed visually but not quantified, electrode wear was not measured gravimetrically, and the low conductivity of their test water may have inflated energy consumption relative to what real, more conductive groundwaters would demand. Future studies, they write, should address long-term operation, pumping energy, electrode consumption, and residue management. Still, the core message stands. Continuous flow electrocoagulation, properly designed with an eye on how water actually flows, can beat the batch systems used to develop it, remove nearly 99 percent of a dangerous contaminant, and do so while cutting energy use by up to 40 percent. For communities sitting atop fluoride-laced aquifers, that is not just an engineering curiosity. It is a credible path to safe water, powered by nothing more exotic than aluminum, electricity, and a well-shaped pipe.

Subject of Research: Batch-to-continuous scale-up of electrocoagulation reactors for fluoride removal from drinking water

Article Title: From batch to continuous electrocoagulation: integrating hydrodynamics and process parameters for efficient fluoride removal and reactor scale-up

Article References: Companhia, B. J. F., & Scalize, P. S. (2026). From batch to continuous electrocoagulation: integrating hydrodynamics and process parameters for efficient fluoride removal and reactor scale-up. Environmental Science and Pollution Research, 33(30), 15329-15351. https://doi.org/10.1007/s11356-026-38203-y

Image Credits: AI Generated

DOI: 10.1007/s11356-026-38203-y

Keywords: electrocoagulation, fluoride removal, water treatment, continuous flow reactor, hydraulic retention time, residence time distribution, axial dispersion, aluminum electrodes, plug flow, energy consumption, reactor scale-up, drinking water

The research team, led by Dayanand Kumbar of the College of Technology and Engineering at MPUAT Udaipur together with P. Shaji James of Kelappaji College of Agricultural Engineering and Technology, Kerala, set out to address a stubborn technical problem. Conventional biogas digesters are poorly suited to high-volume, low-strength liquid wastes like coconut water. These traditional systems demand hydraulic retention times, the length of time liquid spends inside the reactor, of roughly 35 to 55 days. That means enormous digester volumes, high installation costs, and large footprints, all of which push agro-industries toward aerobic treatment systems that consume energy rather than produce it. High-rate anaerobic reactors, by contrast, retain dense microbial populations inside the vessel, allowing retention times to shrink to days or even hours.

The UAHBR tested in this study is a hybrid design that combines two biomass-retention strategies. Wastewater enters at the bottom and flows upward through the reactor. In the lower zone, dense microbial sludge accumulates naturally, while the upper portion is packed with an inert support medium on which biofilms of methane-producing microorganisms can attach and grow. Crucially, the researchers chose a medium that is both cheap and locally abundant: broken coconut shells. The team characterized the shells carefully, measuring a bulk density of 410.95 kilograms per cubic meter, a porosity of 67.56 percent, and a specific surface area of 113.60 square meters per cubic meter, comfortably above the recommended threshold of 100 square meters per cubic meter for biofilm support. The rough, porous surface of the shell fragments encourages microbial attachment, turning an agricultural waste product into a functional component of the energy system itself.

The experimental reactor, fabricated from PVC pipe with an internal diameter of 305 millimeters and a height of 2030 millimeters, held 130 liters of liquid and was installed at the Nila Edible Oil Mill in Kuttippuram. It was seeded with sludge from the existing full-scale UAHBR at Pattambi, filled entirely with effluent, and recirculated for ten days before daily feeding began on the eleventh day. This inoculation strategy proved important: the authors recommend that new systems always start with biomass from a reactor already treating the same or similar effluent, since the microbial community is pre-adapted to the substrate. Once a pseudo-steady state was reached, the researchers systematically shortened the hydraulic retention time from 15 days down to 12, 10, 8, and finally 6 days, tracking how gas production, solids removal, and pH responded to each change.

The results reveal a clear trade-off between how much gas a reactor produces per liter of feed and how much it produces per unit of reactor volume. At the longest retention time of 15 days, the experimental reactor achieved its highest specific biogas production, 225.73 liters per kilogram of total solids added, and a productivity of 8.7 liters of biogas per liter of wastewater treated. Total solids reduction exceeded 80 percent, and biochemical oxygen demand removal reached 84.54 percent at the 12-day retention time, outperforming figures reported for distillery spent wash in earlier studies. As retention times shortened, daily and volumetric gas output rose, peaking at 114 liters per day and 877 liters per cubic meter of reactor volume at the 6-day retention time, but the efficiency of conversion per kilogram of solids declined steadily.

The reason for that decline lies in the chemistry of anaerobic digestion itself. The process unfolds in stages: hydrolysis breaks complex organic molecules apart, acidogenesis converts them into volatile fatty acids, and methanogenic archaea consume those acids to produce methane and carbon dioxide. Coconut water enters the reactor with a pH between roughly 3.2 and 4.9, highly acidic conditions that methanogens tolerate poorly. At longer retention times, the microbial community has time to buffer the incoming acidity, and the effluent pH stabilizes near neutral, between 6.5 and 6.75. When the retention time dropped to 8 days, effluent pH began slipping below 6.3, and at 6 days it fell as low as 5.2, signaling acid accumulation and inhibition of the methane producers. Total solids reduction collapsed from about 80 percent to the 40 to 45 percent range, and biochemical oxygen demand removal fell to just 32 percent. The authors note that beyond a 10-day retention time, performance parameters decline sharply, with the transition from 10 to 6 days showing an exponential rather than linear deterioration.

Perhaps the most striking result came from the full-scale reactor itself. The concrete UAHBR at the Edible Oil Mill in Pattambi, with a total volume of 1.15 cubic meters and the upper 60 percent of its volume packed with coconut shell media, had been operating for more than two years, accumulating a rich biomass inventory. When the researchers standardized its operation, first feeding 60 liters of coconut water daily for a 16.67-day retention time and then reducing to 15 days, the mature system outperformed the experimental unit. At the 15-day retention time it achieved a specific biogas production of up to 354.31 liters per kilogram of total solids added and a productivity of 13.50 liters per liter, with total solids reduction between 79.35 and 81.40 percent. Daily biogas production climbed to roughly 850 liters by the fourth week, all while the influent pH remained below 4, demonstrating that a well-established reactor can neutralize highly acidic feed without chemical pretreatment.

Statistical analysis reinforced the picture. Pearson correlation tests showed strong positive relationships between biogas production and time at the 12-day and 15-day retention times, with correlation coefficients of 0.994 and 0.887 respectively at the 0.01 significance level, while shorter retention times showed negative or no correlation. The researchers also translated the results into practical energy terms. A typical coconut oil mill discharges about 200 liters of waste coconut water per day, which could yield approximately 60 megajoules of energy as biogas containing 60 to 65 percent methane. If burned for thermal applications at 60 percent efficiency, that biogas could replace about 12 kilograms of firewood daily, easing pressure on forests and cutting greenhouse gas emissions. A cost analysis of the system estimated a total installation cost of 310,000 rupees and a payback period of 3.8 years, driven by income from gas, sludge, and treated water.

The study also offers operational guidance for anyone hoping to replicate the system. Start-up should use inoculum from an existing reactor treating similar effluent, begin at a 15-day retention time, and monitor pH closely, neutralizing feed if it drops below pH 4. Loading rates should be changed gradually, since sudden shifts can destabilize the microbial balance. If the goal is to extract maximum energy from every liter of coconut water, retention times of 12 to 15 days are preferable; if maximizing gas output from a smaller reactor is the priority, shorter times around 8 days can be adopted despite lower conversion efficiency. The authors acknowledge remaining challenges, including volatile fatty acid accumulation, long lag phases, and clogging from high solids content, and point to co-digestion, pretreatment technologies, advanced reactor designs, and integrated biorefineries as avenues for improvement, with the nutrient-rich digestate offering potential as a soil fertilizer.

What makes this work notable is its scale and realism. Most anaerobic digestion studies of novel wastewaters remain confined to laboratory beakers, where conditions can be controlled in ways real factories never allow. Here, the reactor ran on genuine mill effluent whose quantity fluctuated with market-driven oil production, and the full-scale system, given time to mature, proved more productive than its carefully managed experimental counterpart. For the millions of farming families across Kerala, Tamil Nadu, Karnataka, and Andhra Pradesh who depend on the coconut palm, and for the kopra and oil mills that process its harvest, the message is that the wastewater they currently discard is not waste at all. It is a feedstock, and with the right biology and the right reactor, it can close the loop between agro-industrial pollution and renewable energy in a genuinely circular economy.

Subject of Research: Biogas production from waste coconut water using a full-scale upflow anaerobic hybrid bioreactor

Article Title: Investigation on biogas production from coconut water using a full-scale upflow anaerobic hybrid bioreactor

Article References: Kumbar, D., James, P. S., Kumar, V. K., Hallad, S. C., & Ramappa, D. (2026). Investigation on biogas production from coconut water using a full-scale upflow anaerobic hybrid bioreactor. Discover Green Chemistry, 1(1), Article 2. https://doi.org/10.1007/s44509-026-00002-6

Image Credits: AI Generated

DOI: 10.1007/s44509-026-00002-6

Keywords: biogas, coconut water, anaerobic digestion, upflow anaerobic hybrid bioreactor, hydraulic retention time, wastewater treatment, renewable energy, circular economy, Kerala, agro-industrial waste, methanogenesis, coconut shell media