The Scale of Contaminated Water at Fukushima

To understand the scale of the treatment challenge, it helps to look at the numbers. As of early 2025, more than 1.3 million cubic meters of treated water were stored in over a thousand large tanks across the Fukushima Daiichi site. That volume is still increasing by roughly 100 to 130 cubic meters per day, primarily from groundwater inflow that seeps into the damaged reactor buildings and mixes with cooling water continuously injected to keep the melted fuel subcooled. The accumulation is slowed by an extensive system of groundwater bypass wells, sub-drains, and an impermeable frozen-soil wall, but it cannot be completely stopped. Every liter of stored water needs to be treated to a level that allows either safe controlled release or long-term storage. The sheer quantity dictated that treatment technologies be not only effective but also scalable, robust, and capable of handling a mixture of over 60 different radionuclides.

Radioactive water at Fukushima is categorized by the concentration and types of isotopes it contains. Freshly pumped reactor building water can contain high levels of cesium-134, cesium-137, strontium-90, iodine-129, and a range of other fission and activation products, including the difficult-to-remove isotope tritium. The initial emergency cooling phase saw crude decontamination attempts using zeolite sandbags and simple filtration, but it quickly became clear that more systematic multi-stage treatment was needed to reduce risks and eventually free up storage space. The response was the development and deployment of a suite of innovative, layered treatment technologies, each targeting specific radionuclide groups or physical characteristics of the wastewater. The design philosophy evolved from emergency first-response systems into a permanent, continuously optimized processing train capable of handling variable influent chemistry.

Primary Treatment Systems Deployed

The water treatment process at Fukushima Daiichi is not a single machine but an integrated chain of systems, each with a distinct role. Early efforts centered on emergency cesium removal, followed by broader multi-nuclide separation, and then by polishing and waste solidification. The main facilities that have come to define this effort are the Kurion and SARRY cesium adsorption systems, the Advanced Liquid Processing System (ALPS), various reverse osmosis (RO) membrane units, and long-term vitrification trials. Each system uses fundamentally different physical and chemical principles to achieve radionuclide separation. The overall architecture is designed for redundancy: if one stage fails or requires maintenance, the others can continue operating, and bypass lines allow water to be redirected for re-treatment if initial results exceed regulatory limits.

Cesium Removal: Kurion and SARRY Adsorption Systems

The first priority after stabilizing reactor cooling was to remove the isotopes responsible for the bulk of external radiation dose: cesium-134 and cesium-137. Early circulation cooling systems included zeolite-based adsorption vessels, but these quickly became overwhelmed due to the high salinity of seawater-contaminated water and the sheer volume of flow. Two larger-scale, mobile cesium removal systems were rapidly deployed: the Kurion system, which used a proprietary inorganic ion-exchange media based on ammonium molybdophosphate and multi-stage columns for cesium and strontium, and the Simplified Active Water Retrieve and Recovery System (SARRY), which employed a different type of granular inorganic adsorbent, primarily sodium titanate and crystalline silicotitanate, optimized for high-salinity water. Both systems were designed for rapid deployment and modular expansion, allowing operators to increase capacity as the volume of stored water grew.

The Kurion system processed contaminated water at rates of up to 1,200 cubic meters per day during its peak operation. The SARRY system, designed and built by Toshiba, could handle 500 cubic meters per day per train, with multiple trains operating in parallel. Both systems demonstrated over a 99.9% reduction in cesium activity, bringing water from millions of becquerels per liter down to levels where it could be sent to secondary treatment without overwhelming downstream equipment. The solid adsorbents loaded with high-activity cesium became a secondary waste stream, eventually destined for vitrification or cementation. Operational challenges included adsorbent fouling from organic matter and biofilms, which required periodic chemical cleaning and media replacement. By 2013, the combination of Kurion and SARRY had processed over 400,000 cubic meters of water, demonstrating the viability of large-scale inorganic ion exchange under accident conditions.

Advanced Liquid Processing System (ALPS)

The centerpiece of the multi-nuclide removal strategy is the ALPS, often described as the most sophisticated radionuclide removal system ever deployed at a nuclear accident site. Unlike the initial cesium-only systems, ALPS is designed to strip out 62 of the 63 radionuclides listed in Japanese regulatory standards, excluding only tritium. The system, developed by Toshiba and partners, comprises a train of chemically distinct column reactors, each optimized for a different group of isotopes. The design went through multiple generations, with ALPS-II and ALPS-III incorporating lessons learned from early operational data, including improved pretreatment to prevent precipitates from clogging the ion-exchange columns.

ALPS works in three primary stages. In the first, iron co-precipitation is performed by adding ferric sulfate and adjusting pH to around 8–9, which helps scavenge a variety of metal radionuclides such as cobalt, manganese, and ruthenium isotopes, forming a settleable iron hydroxide floc that is filtered out using plate-and-frame filters or centrifuges. The second stage uses a series of highly selective ion-exchange resins and inorganic adsorbents. One column targets strontium, barium, and radium with a proprietary crystalline material (a titanate-based adsorbent); another removes alpha-emitting transuranic elements (americium, curium, plutonium) using an engineered extractant impregnated on a porous support; yet another captures selenium and antimony using a chelating resin. The final stage involves a "multi-nuclide removal column" containing a mixed bed of high-performance adsorbents, including activated carbon for iodine and silver-impregnated zeolites, to polish residual traces of the broad spectrum. The system processes up to 500 cubic meters per day per train, and multiple trains have been constructed and improved over the years, with current operations using six trains in parallel.

The output water from ALPS, known as "ALPS-treated water," meets the discharge concentration limits for all radionuclides except tritium. However, early batches of ALPS-treated water were found to have residual radionuclides above regulatory limits due to initial performance issues and the enormous variability of the incoming waste stream. This necessitated re-treatment through upgraded ALPS systems with enhanced adsorbents and additional selective columns, a process TEPCO and the Japanese government have verified through extensive sampling campaigns. The monitoring data for current treated water confirm that, after secondary ALPS polishing, concentrations of key isotopes such as Cs-137, Sr-90, and I-129 are well below the World Health Organization drinking-water guidance levels. The system's reliability has improved dramatically since 2015, with over 99.9% removal efficiency consistently achieved across all regulated nuclides.

Membrane Filtration and Reverse Osmosis

Membrane technologies have played a dual role in Fukushima water management: desalination and final purification. Contaminated water at the site is highly saline because seawater was used for emergency cooling in the earliest days, and groundwater intrusion continues to bring dissolved salts. High salt content interferes with certain ion-exchange processes and increases the total volume of waste if discharged. Reverse osmosis (RO) units were thus integrated both before and after ALPS treatment.

Pre-ALPS, an RO system removes a large fraction of dissolved salts, reducing the load on downstream adsorbents and minimizing background competition for ion-exchange sites. Typical reduction in total dissolved solids (TDS) is from 30,000–35,000 mg/L down to below 500 mg/L, turning seawater into near-fresh water. This step also removes colloidal particles and some dissolved organic compounds that could foul the ion-exchange resins. Post-ALPS, a second RO step can further polish the water, concentrating any residual trace nuclides and producing clean permeate that, after verification, could be released under stringent controls or reused as cooling water. RO membranes with very high rejection rates — often thin-film composite polyamide membranes with salt rejection above 99.5% — are used. The concentrate stream (brine) containing a minor fraction of radionuclides and all the rejected salts is then sent back for re-treatment or solidification. Energy consumption and membrane fouling remain operational challenges, but careful chemical pretreatment with antiscalants and scheduled cleaning cycles keep the systems running reliably.

In addition to RO, ultrafiltration and nanofiltration are occasionally employed to remove suspended colloidal particles that might carry sorbed radionuclides, providing an additional barrier before final discharge. Together, membrane stages act as robust physical barriers that complement the chemical selectivity of ion-exchange systems. The Fukushima facility operates one of the largest accident-site RO installations globally, with a total capacity exceeding 2,000 cubic meters per day.

Ion Exchange Resins for Targeted Radionuclide Removal

Beyond the large integrated systems, a variety of specialized ion exchange resins and materials are used to capture specific problematic isotopes. For example, silver-impregnated activated carbon and silver-exchanged zeolites are extremely effective at removing iodine-129, a long-lived fission product (half-life 15.7 million years) that can accumulate in the human thyroid. Barium sulfate co-precipitation followed by selective cation exchange using a barium-specific resin is used for radium removal. Crystalline silicotitanate (CST), developed originally for defense waste at the Savannah River Site, has been proven effective for capturing both cesium and strontium from high-pH, high-salt solutions and is studied as a potential future polishing material at Fukushima. Chitosan-based and other biosorbents have also been investigated for their ability to chelate certain radionuclides, though industrial-scale deployment has been limited due to lower chemical stability in high-radiation fields.

The effective use of ion exchange relies on understanding the solution chemistry. Competing ions like sodium, calcium, and magnesium can occupy exchange sites, reducing capacity for target radionuclides. This is why pretreatment steps like RO and pH adjustment are so important. Research published by the Japan Atomic Energy Agency (JAEA) and other institutions has detailed how the selectivity of CST and titanate-based media can be optimized, and new composite materials are being developed that combine ion-exchange functionality with structural robustness for continuous column operation. For instance, researchers have grafted titanate nanofibers onto macroporous polymer beads, achieving both high capacity and low pressure drop in column operations. Such innovations promise to reduce waste volumes further by increasing the loading of radionuclides per unit mass of adsorbent.

Vitrification: Stabilizing High-Activity Waste

All the treatment processes generate secondary radioactive waste: spent adsorption columns, sludges, and concentrates that contain a highly concentrated mix of radionuclides. In line with international best practices for high-level waste, Japan plans to immobilize this waste by vitrification — converting it into a durable glass form. Vitrification involves mixing the waste with glass-forming chemicals (such as silica, borax, and other fluxes) at temperatures exceeding 1,100°C and pouring the melt into stainless steel canisters. The resulting borosilicate glass matrix chemically bonds the radionuclides, making them highly resistant to leaching even if the canister corrodes over geological time scales.

A pilot vitrification facility is in operation at Fukushima, intended to process the cesium- and strontium-loaded adsorbents into glass logs. Early tests encountered difficulties with uneven heating and volatile radionuclide release, particularly cesium volatility at high temperatures. Iterative design improvements — including the addition of cold-cap layers, off-gas treatment systems, and advanced melters with induction heating — have led to a more reliable process. The vitrification product is destined for interim storage and eventual deep geological disposal, likely at a future Japanese repository. While the cost and complexity of vitrification are high, it provides the most stable and volume-efficient waste form currently available for the long-lived isotopes isolated from the water treatment streams. Alternative stabilization methods, such as cementation or geopolymer encapsulation, are also being investigated for lower-activity waste streams to reduce overall costs.

The Tritium Challenge and Ongoing Research

Tritium, a radioactive isotope of hydrogen with a half-life of 12.3 years, is the one radionuclide that conventional water treatment cannot remove at scale. Because tritium replaces ordinary hydrogen in water molecules to form tritiated water (HTO), it behaves chemically identically to regular water, making separation via filtration, ion exchange, or precipitation extraordinarily difficult. Yet tritium at Fukushima is present in ALPS-treated water at concentrations around 1.5 million becquerels per liter, which must be reduced before release to meet Japan's regulatory discharge limit of 60,000 Bq/L — and in practice, TEPCO dilutes the treated water with seawater to bring the tritium concentration down to 1,500 Bq/L, far below the WHO drinking-water guideline of 10,000 Bq/L.

While large-scale tritium removal remains commercially unavailable, several innovative laboratory and pilot-scale technologies are under investigation. These include electrolytic enrichment (where tritium accumulates in the electrolyte due to kinetic isotope effects), chemical exchange (the Girdler sulfide process or combined electrolysis catalytic exchange, CECE), cryogenic distillation of water or hydrogen, and membrane distillation enhanced with isotopic selectivity using materials like graphene oxide. Molecular sieving using graphene oxide membranes or nanoporous metal-organic frameworks has shown some promise in selectively impeding tritiated water transport due to differences in molecular size and hydrogen bonding, but throughput and durability issues persist. Another approach is to separate tritium from the water after converting it to tritium gas via a catalytic exchange reaction, but the energy costs are extremely high, making it impractical for massive volumes. Until a breakthrough occurs, dilution and monitored discharge remain the pragmatic strategy, aligned with the practice of nuclear facilities worldwide. The International Atomic Energy Agency has endorsed this approach, noting that it is consistent with radiological protection principles and discharge practices at other coastal nuclear sites.

Safety, Monitoring, and Controlled Discharge

In August 2023, Japan began the controlled release of ALPS-treated water into the Pacific Ocean, a decision endorsed by the IAEA after a comprehensive review that included expert peer reviews of safety assessments, operational readiness, and environmental monitoring plans. The discharge process involves re-purifying the water through secondary ALPS treatment to ensure all radionuclides other than tritium are below regulatory limits (with a margin of error), then blending it with large volumes of seawater in a dilution pit to bring tritium down to the 1,500 Bq/L target. The diluted water is then released through an undersea tunnel one kilometer offshore, where it disperses rapidly due to ocean currents and turbulence. The release is staged over 30 years to match the rate of water treatment and storage, and the total tritium discharged per year is less than that released by many coastal nuclear power plants under normal operation.

A rigorous monitoring program — involving TEPCO, the Japanese Nuclear Regulation Authority, and the IAEA — continuously measures radioactivity in the discharged water, the seawater near the outfall, marine sediments, and biota such as fish, shellfish, and seaweed. Real-time data is published online through open-access platforms. According to IAEA reports, the tritium levels in fish samples from the Fukushima region have remained well within Japan's food safety standards (below 1,000 Bq/kg for tritium-free water tritium, which is far below the limit of 60,000 Bq/kg set for tritium), and there has been no measurable increase in the already low background tritium levels in the surrounding sea. Independent monitoring by third-party laboratories and scientific institutions confirms these findings. The multi-barrier treatment approach, combined with dilution and environmental surveillance, provides a robust framework that sets a benchmark for future nuclear legacy cleanup projects. The Nuclear Regulation Authority Japan publishes regular updates on the monitoring results.

Future Innovations and Long-Term Sustainability

The Fukushima cleanup has already stimulated a wave of research into more efficient, lower-waste treatment processes. Advanced nanomaterials, such as functionalized carbon nanotubes and graphene oxide frameworks, are being studied for their high specific surface area and tunable selectivity against strontium and cesium in saline solutions. For example, researchers at the Japan Atomic Energy Agency have developed a composite of magnetic nanoparticles coated with titanate that can be recovered using a magnetic field after adsorption, eliminating the need for column operations and reducing secondary waste. Biomimetic approaches, including engineered proteins and peptides that selectively bind radionuclides, are in early-stage development, aiming for ultra-high selectivity at low concentrations. Some research groups are exploring the use of electrochemical separation techniques, where an applied electric field drives specific ions into a collection stream, potentially avoiding the generation of large volumes of spent adsorbent — this approach has shown 80–90% removal efficiency for cesium in pilot tests using carbon cloth electrodes.

Machine learning and digital twin modeling are also being applied to optimize the operation of existing treatment trains. By predicting breakthrough curves for ion-exchange columns using historical data and real-time sensors, operators can adjust flow rates, regeneration cycles, and column switching schedules to maximize media lifetime and minimize re-treated water volumes. This predictive maintenance approach has already reduced operating costs by an estimated 15–20% in trial runs. The goal over the next decade is not just to process the current inventory of stored water (estimated at 1.35 million cubic meters), but to create a closed-loop system where reactor cooling water is continuously treated and reused, dramatically reducing the need for new water intake and limiting the creation of additional contaminated water. A demonstration loop is planned for 2026, incorporating ALPS, RO, and tritium separation using a CECE pilot unit. The knowledge gained from Fukushima is already informing preparedness strategies for other nuclear sites facing similar water management challenges, including decommissioning projects at Chernobyl, Sellafield, and Hanford.

Conclusion

The Fukushima Daiichi cleanup has served as a crucible for innovation in water treatment technology. The integration of cesium-adsorption systems, ALPS multi-nuclide removal, membrane filtration, and advanced vitrification demonstrates how a layered treatment architecture can address an extraordinarily complex radioactive waste stream. While the tritium challenge remains, the combination of proven separation technologies, rigorous monitoring, and ongoing R&D ensures that the project continues to meet safety, environmental, and regulatory milestones. The lessons learned — in process intensification, waste minimization, and transparent scientific communication — will shape the future of nuclear environmental remediation worldwide. As the Japan Atomic Energy Agency and TEPCO continue to refine these technologies, the Fukushima experience provides a powerful template for managing liquid radioactive waste at scale in a politically and scientifically rigorous manner.

References and Further Reading
• Tokyo Electric Power Company Holdings (TEPCO) – ALPS Treated Water Portal
• International Atomic Energy Agency – Fukushima Daiichi Status Updates
• Japan Atomic Energy Agency – R&D on Decommissioning
• Nuclear Regulation Authority Japan – Monitoring Information