Technology brief
What this platform addresses
In the last week of July 2026 the flow of the Danube fell to approximately 1,630 m³/s, and in the first days of August to approximately 1,400 m³/s, against a long-term August mean near 3,900 m³/s and an…
Energy Production
In the last week of July 2026 the flow of the Danube fell to approximately 1,630 m³/s, and in the first days of August to approximately 1,400 m³/s, against a long-term August mean near 3,900 m³/s and an…
Technology brief
In the last week of July 2026 the flow of the Danube fell to approximately 1,630 m³/s, and in the first days of August to approximately 1,400 m³/s, against a long-term August mean near 3,900 m³/s and an…
The challenge
Nuclear and thermal power generation. Retrofit of once-through condenser circuits at plants whose permitted output is constrained by river temperature or flow — the case documented for the 2026 Danube events and the approved forward pipeline of new once-through units on the same river.
Combined-cycle and gas-fired stations, refineries, metallurgical and chemical plants, pulp and paper mills, large compressor stations. Same condenser/process-cooling duty and the same once-through failure mode outside the nuclear sector.
Data centers, LNG terminals, desalination plants, district cooling systems. Heat-rejection duty where water scarcity or discharge-temperature limits constrain capacity.
Hospitals and hotels. Facilities under strict aerosol and Legionella liability regulation where an open evaporative basin is a compliance risk.
Users: nuclear and thermal plant operators, refinery and metallurgical operators, data center operators, desalination plant operators, municipal district-cooling utilities.
ARBOK solution
In the last week of July 2026 the flow of the Danube fell to approximately 1,630 m³/s, and in the first days of August to approximately 1,400 m³/s, against a long-term August mean near 3,900 m³/s and an annual norm of roughly 3,000 m³/s. Romania placed a unit of a two-unit CANDU nuclear station into a controlled shutdown to protect its cooling pumps, kept the second unit online after case-by-case analysis, and undertook emergency river works — removal of a rock obstruction with approximately 180 kg of explosives and deliberate sinking of loaded barges to divert flow into the channel feeding the plant. In Hungary, a station that supplies roughly half of national generation ran its four units at a fraction of combined capacity. Both episodes were reported publicly as hydrological events — low river, therefore the plant stops. ARBOK-PowerCool reframes them as thermal rejection failures: a once-through condenser circuit cannot dispose of its waste heat when the receiving river is simultaneously low and warm, and the binding constraint is the thermal power that must leave the site every hour, not the electrical output of the turbines.
ARBOK-PowerCool is assembled from standardized containerized modules and installed in one of two configurations: at the intake, holding a set cold line ahead of the existing condenser with no change to plant design (Mode A); or as a conversion of the condenser circuit to a closed, once-charged loop that removes river abstraction altogether (Mode B). The working fluid is ordinary water; the process runs at ambient temperature with no external heat input, no heating circuit and no process-temperature setpoint — water boils under deep vacuum and, in evaporating, extracts latent heat from the process stream. The vapor is then condensed within a closed loop, and up to 98% of the transferred heat is returned into the cycle rather than discharged to atmosphere as in a wet tower. Unlike a wet cooling tower, where a larger temperature difference means more evaporation and a larger makeup water bill, in this cycle a larger delta means more energy carried as vapor and therefore higher absolute internal recovery — a hotter inlet improves rather than degrades relative efficiency, which is the property that matters in a heatwave. The same module family cleans the circulating water of salts, oils, radionuclides and dust in the same pass, with no membranes, no catalysts and no liquid tailings.
The operating principle in every module is phase transition under deep vacuum. The working fluid is ordinary water. In a low-pressure zone, water boils at ambient temperature and, in evaporating, extracts latent heat from the process stream. No heat is supplied from outside the system; there is no heating circuit and no process-temperature setpoint — the temperature of the process equals the temperature of the incoming stream and of the surroundings. The vapor is then condensed within a closed loop, and up to 98% of the transferred heat is returned into the cycle, because vapor enthalpy exceeds that of the liquid, so condensation feeds the next evaporation stage rather than discharging to atmosphere. There is no compressor, no refrigerant, no waterfall, no forced draft, no plume, and no open contact between the working water and the air.
The behavior of the cycle with respect to temperature difference is the inverse of the classical case: in a wet tower, a larger delta means more evaporation and a larger makeup bill; here a larger delta means more energy carried as vapor and therefore higher absolute recovery, so a hotter inlet improves relative efficiency instead of degrading it.
Water quality is handled in the same pass rather than in a separate plant. The circulating water is cleaned of salts, oils, radionuclides and dust, with no membranes, no catalysts and no consumable chemistry. Impurities leave in dry form as a solid residue; the material balance is closed and there are no liquid tailings. A secondary-heat module, TERU, recovers parasitic heat across the temperature differences of the installation and returns it as electricity, reducing grid draw. Instrumentation is continuous: ARBOK-Sonar measures water parameters around the clock, reports to the control center, and on deviation either alerts operators or moves the installation into a protective mode up to shutdown.
Market and application
The addressable market for cooling and treatment of thermal and nuclear station water is estimated at $50+ billion per year worldwide, breaking down by share of global thermal generation: approximately $15 billion per year in China, the largest thermal fleet; $7–8 billion per year in the United States, with its once-through reactor fleet and large steam fleet; approximately $2 billion per year in France, Europe's most acute river-cooling case. With refineries, metallurgy, chemicals, data centers and desalination included, the addressable figure is several times larger.
Geography is the entire heat belt: France, Switzerland, Germany, Spain, Italy, Romania, Hungary, Turkey, the United States, India, China and the Middle East. The forward pipeline repeats the same architecture that failed in 2026: further once-through nuclear capacity on the Danube has passed hearings and received national regulatory approval, adding a further 2 × 1,500 MWth of once-through rejection duty on the same river, alongside a second approved project comprising two additional reactors designed for direct water cooling from the Danube. This approved pipeline is itself a forward market signal.
Regulatory direction of travel supports the category: inland nuclear stations in Russia are required to install cooling towers, with only coastal stations exempted; Uzbekistan has moved toward dry-tower technology after consultation. The principle that inland thermal generation must stop treating rivers as heat sinks is already established in regulatory practice; what has been missing is a rejection path that neither consumes water nor loses capability in the heat.
Reference node used throughout the ARBOK cooling series: 10,000 m³/h of circulation at 8,000 operating hours per year, permitting direct comparison with wet-tower practice.
Wet tower (classical), annual cost: electricity for pumps and fans, $576,000–672,000; makeup water at 3.8% of circulation and $1.50 per m³, approximately $4,560,000 at a delta near 16 °C, rising to approximately $10,440,000 at a delta of 35–37 °C; chemistry, biocides, laboratory work, cleaning and repair, $160,000–480,000. Total: approximately $5,296,000–5,712,000 per year at the lower delta and $11,176,000–11,592,000 per year at the higher one.
Closed vacuum cycle, annual cost: the water line is zero, because water is charged once and circulates without loss. Electricity after the TERU reduction, $230,000–384,000; planned service, $200,000–300,000. Total: approximately $430,000–684,000 per year, and it does not inflate with delta.
Capital cost: a comparable classical node — tower, basin, pumping station and installation — commonly budgeted at up to $10,000,000. Modular container equipment for the same duty: $4,000,000–6,000,000, a capital advantage of $4,000,000–6,000,000 at the outset. Payback is about 7 years — the standard horizon for industrial cooling assets — against operating savings of approximately $4,612,000–5,282,000 per year at the lower delta; at the higher delta, savings exceed $10,000,000 per year.
Avoided-derating value (site-specific, larger than the operating-cost terms above). A 1.3 GW unit halved for one week of heat forgoes over 100,000 MWh — 20–40 million € of foregone revenue in a single week at the 200–400+ €/MWh prices that prevail in those hours, repeated several times per season; an operator who has sold capacity forward must additionally buy replacement energy on the spot market at the same peak prices. Fleet-wide, river-condition curtailment is estimated at approximately 0.5 TWh per year for one national nuclear fleet — roughly 0.3% of its annual generation — worth on the order of 200 million € per year at 200 €/MWh.
Use cases
Nuclear and thermal power generation. Retrofit of once-through condenser circuits at plants whose permitted output is constrained by river temperature or flow — the case documented for the 2026 Danube events and the approved forward pipeline of new once-through units on the same river.
Combined-cycle and gas-fired stations, refineries, metallurgical and chemical plants, pulp and paper mills, large compressor stations. Same condenser/process-cooling duty and the same once-through failure mode outside the nuclear sector.
Data centers, LNG terminals, desalination plants, district cooling systems. Heat-rejection duty where water scarcity or discharge-temperature limits constrain capacity.
Hospitals and hotels. Facilities under strict aerosol and Legionella liability regulation where an open evaporative basin is a compliance risk.
Users: nuclear and thermal plant operators, refinery and metallurgical operators, data center operators, desalination plant operators, municipal district-cooling utilities.
Mode A — intake pre-cooling (the bridge). Modules are installed on the abstraction side and hold the intake stream at a set cold line. No change is made to the condenser, the circulating pumps, the discharge structure, or any component inside the licensed process boundary — the station is handed colder water than the source is providing, rather than being asked to accept a new cooling system. This mode buys an operating season without a licensing program and is reversible.
Mode B — closed circulating loop. The condenser circuit is charged with water once and then run as a closed working medium: no discharge, no continuous makeup, no open contact with air. The abstraction requirement disappears, and station availability is decoupled from river stage and river temperature entirely.
The two modes form a sequence rather than a choice: Mode A first, where the timeline is measured in a single construction season; Mode B where the site and the regulator permit a fuller conversion. Operation in both modes is continuously monitored by ARBOK-Sonar, with automatic alerts or protective shutdown on deviation from set water-quality parameters.
ARBOK-Sonar for continuous water-quality monitoring and automatic protective shutdown. TERU for secondary-heat recovery, converting parasitic heat back into electricity and reducing the installation's own grid draw. Shares its 50–2,000+ MWth / 20–200 MWth modular thermal envelope with ARBOK-CoolTower, the closed-loop cooling-tower replacement platform; PowerCool applies the same phase-change and closed-loop principles specifically as a dual-mode retrofit for existing once-through condenser circuits, and references the same closed-loop economics developed for ARBOK-ClosedTower. The underlying phase-change principle is demonstrated at smaller scale on Arbok-Criojet (vacuum-nozzle cooling, TRL 5), whose CRIOJET performance data is cited directly as the physics validation for PowerCool's circulating-water envelope.
In the last week of July 2026 the flow of the Danube fell to approximately 1,630 m³/s, and in the first days of August to approximately 1,400 m³/s, against a long-term August mean near 3,900 m³/s and an annual norm of roughly 3,000 m³/s. Romania placed a unit of a two-unit CANDU nuclear station into a controlled shutdown to protect its cooling pumps, kept the second unit online after case-by-case analysis, and undertook emergency river works — removal of a rock obstruction with approximately 180 kg of explosives and deliberate sinking of loaded barges to divert flow into the channel feeding the plant. In Hungary, a station that supplies roughly half of national generation ran its four units at a fraction of combined capacity. Both episodes were reported publicly as hydrological events — low river, therefore the plant stops. ARBOK-PowerCool reframes them as thermal rejection failures: a once-through condenser circuit cannot dispose of its waste heat when the receiving river is simultaneously low and warm, and the binding constraint is the thermal power that must leave the site every hour, not the electrical output of the turbines.
ARBOK-PowerCool is assembled from standardized containerized modules and installed in one of two configurations: at the intake, holding a set cold line ahead of the existing condenser with no change to plant design (Mode A); or as a conversion of the condenser circuit to a closed, once-charged loop that removes river abstraction altogether (Mode B). The working fluid is ordinary water; the process runs at ambient temperature with no external heat input, no heating circuit and no process-temperature setpoint — water boils under deep vacuum and, in evaporating, extracts latent heat from the process stream. The vapor is then condensed within a closed loop, and up to 98% of the transferred heat is returned into the cycle rather than discharged to atmosphere as in a wet tower. Unlike a wet cooling tower, where a larger temperature difference means more evaporation and a larger makeup water bill, in this cycle a larger delta means more energy carried as vapor and therefore higher absolute internal recovery — a hotter inlet improves rather than degrades relative efficiency, which is the property that matters in a heatwave. The same module family cleans the circulating water of salts, oils, radionuclides and dust in the same pass, with no membranes, no catalysts and no liquid tailings.
Nuclear and thermal power generation. Retrofit of once-through condenser circuits at plants whose permitted output is constrained by river temperature or flow — the case documented for the 2026 Danube events and the approved forward pipeline of new once-through units on the same river.
Combined-cycle and gas-fired stations, refineries, metallurgical and chemical plants, pulp and paper mills, large compressor stations. Same condenser/process-cooling duty and the same once-through failure mode outside the nuclear sector.
Data centers, LNG terminals, desalination plants, district cooling systems. Heat-rejection duty where water scarcity or discharge-temperature limits constrain capacity.
Hospitals and hotels. Facilities under strict aerosol and Legionella liability regulation where an open evaporative basin is a compliance risk.
Users: nuclear and thermal plant operators, refinery and metallurgical operators, data center operators, desalination plant operators, municipal district-cooling utilities.
The operating principle in every module is phase transition under deep vacuum. The working fluid is ordinary water. In a low-pressure zone, water boils at ambient temperature and, in evaporating, extracts latent heat from the process stream. No heat is supplied from outside the system; there is no heating circuit and no process-temperature setpoint — the temperature of the process equals the temperature of the incoming stream and of the surroundings. The vapor is then condensed within a closed loop, and up to 98% of the transferred heat is returned into the cycle, because vapor enthalpy exceeds that of the liquid, so condensation feeds the next evaporation stage rather than discharging to atmosphere. There is no compressor, no refrigerant, no waterfall, no forced draft, no plume, and no open contact between the working water and the air.
The behavior of the cycle with respect to temperature difference is the inverse of the classical case: in a wet tower, a larger delta means more evaporation and a larger makeup bill; here a larger delta means more energy carried as vapor and therefore higher absolute recovery, so a hotter inlet improves relative efficiency instead of degrading it.
Water quality is handled in the same pass rather than in a separate plant. The circulating water is cleaned of salts, oils, radionuclides and dust, with no membranes, no catalysts and no consumable chemistry. Impurities leave in dry form as a solid residue; the material balance is closed and there are no liquid tailings. A secondary-heat module, TERU, recovers parasitic heat across the temperature differences of the installation and returns it as electricity, reducing grid draw. Instrumentation is continuous: ARBOK-Sonar measures water parameters around the clock, reports to the control center, and on deviation either alerts operators or moves the installation into a protective mode up to shutdown.
Vacuum: deep vacuum. Process temperature: ambient, no external heat, no setpoint. Internal heat recovery: up to 98%. Water recovery: up to 99.98%, closed material balance, no liquid tailings. Standard module: containerized, shippable format, scaled by module count. Thermal envelope: 50–2,000+ MWth, in modules of 20–200 MWth. Water treatment energy: 0.5–1 kWh per m³ of processed flow (preprint figure) — reported elsewhere in the source material as 2–3 kWh/t for the same combined cooling-and-treatment duty (ARBOK-PowerCool article). TERU secondary-heat recovery: reduces grid draw by up to 20% (preprint) — reported elsewhere as up to a quarter (ARBOK-PowerCool article and post).
Physics validation: the same phase-change principle is demonstrated at module scale on the CRIOJET line — a large water volume taken through a wide temperature delta within minutes on a small energy input (see Arbok-Criojet).
Case example — Danube retrofit, 2026:
| Parameter | Value |
|---|---|
| Reactor thermal power per unit | ≈2,180 MWth |
| Electrical output per unit | ≈700 MWe |
| Site rejection duty (two units) | ≈3,000 MWth |
| Design intake flow | 108 m³/s (54 m³/s per unit) |
| Temperature rise at design flow | ≈6.6 °C |
| Regulatory discharge ceiling | 28 °C |
| Compliant-intake threshold | ≈21.4 °C |
| Mode A duty for a 4 K intake reduction | ≈1,809 MWth (≈9–90 modules of the 20–200 MWth class) |
| France, single curtailment event, 13 July 2026 | 6.4 GW cut, ≈14% of demand |
Standardized containerized modules, mounted on prepared ground without high civil works, scaled by module count. Core stages: vacuum evaporation stage; condenser/heat-recovery loop returning up to 98% of transferred heat internally; TERU secondary-heat module converting recovered parasitic heat into electricity; integrated water-treatment stage separating salts, oils, radionuclides and dust into dry fractions; ARBOK-Sonar continuous monitoring with automatic protective mode. Two field configurations built on the same module: Mode A, intake pre-cooling modules installed ahead of the existing condenser with no modification to the condenser, circulating pumps or discharge structure; Mode B, the same modules charging and running the condenser circuit as a closed loop, removing the river abstraction requirement.
Efficiency rises with delta, not against it. The inverse of the classical wet-tower relationship — a hotter inlet, the exact condition of a heatwave, improves the unit's relative recovery instead of degrading it.
No modification to the licensed process boundary in Mode A. Intake pre-cooling restores compliance without any change inside a plant's existing design or safety case, which is the constraint that renders conventional retrofits useless on the timescale of a heatwave.
Installable in months, not years. Containerized, modular, no high civil works — in contrast with redesign of an operating plant's cooling architecture, which is a multi-year regulatory and engineering program.
Zero water consumption in Mode B. The abstraction requirement disappears entirely once the condenser circuit runs as a closed, once-charged loop.
No open basin, no plume, no Legionella habitat. No contact between the working water and open air at any stage.
Reversibility. Mode A is explicitly a bridge — it can be removed or extended without committing the site to Mode B.
ARBOK-Sonar for continuous water-quality monitoring and automatic protective shutdown. TERU for secondary-heat recovery, converting parasitic heat back into electricity and reducing the installation's own grid draw. Shares its 50–2,000+ MWth / 20–200 MWth modular thermal envelope with ARBOK-CoolTower, the closed-loop cooling-tower replacement platform; PowerCool applies the same phase-change and closed-loop principles specifically as a dual-mode retrofit for existing once-through condenser circuits, and references the same closed-loop economics developed for ARBOK-ClosedTower. The underlying phase-change principle is demonstrated at smaller scale on Arbok-Criojet (vacuum-nozzle cooling, TRL 5), whose CRIOJET performance data is cited directly as the physics validation for PowerCool's circulating-water envelope.
Mode A — intake pre-cooling (the bridge). Modules are installed on the abstraction side and hold the intake stream at a set cold line. No change is made to the condenser, the circulating pumps, the discharge structure, or any component inside the licensed process boundary — the station is handed colder water than the source is providing, rather than being asked to accept a new cooling system. This mode buys an operating season without a licensing program and is reversible.
Mode B — closed circulating loop. The condenser circuit is charged with water once and then run as a closed working medium: no discharge, no continuous makeup, no open contact with air. The abstraction requirement disappears, and station availability is decoupled from river stage and river temperature entirely.
The two modes form a sequence rather than a choice: Mode A first, where the timeline is measured in a single construction season; Mode B where the site and the regulator permit a fuller conversion. Operation in both modes is continuously monitored by ARBOK-Sonar, with automatic alerts or protective shutdown on deviation from set water-quality parameters.
The phase-change principle underlying ARBOK-PowerCool is validated at module scale on the CRIOJET line (see Arbok-Criojet), assessed there at TRL 5 (confirmed by Michael). ARBOK-PowerCool scales the same principle into a larger thermal envelope for circulating-water duty at power stations and comparable industrial sites; deployment-specific validation evidence at station scale, and a PowerCool-specific TRL figure independent of the CRIOJET analogy, [требует уточнения из базы].
The addressable market for cooling and treatment of thermal and nuclear station water is estimated at $50+ billion per year worldwide, breaking down by share of global thermal generation: approximately $15 billion per year in China, the largest thermal fleet; $7–8 billion per year in the United States, with its once-through reactor fleet and large steam fleet; approximately $2 billion per year in France, Europe's most acute river-cooling case. With refineries, metallurgy, chemicals, data centers and desalination included, the addressable figure is several times larger.
Geography is the entire heat belt: France, Switzerland, Germany, Spain, Italy, Romania, Hungary, Turkey, the United States, India, China and the Middle East. The forward pipeline repeats the same architecture that failed in 2026: further once-through nuclear capacity on the Danube has passed hearings and received national regulatory approval, adding a further 2 × 1,500 MWth of once-through rejection duty on the same river, alongside a second approved project comprising two additional reactors designed for direct water cooling from the Danube. This approved pipeline is itself a forward market signal.
Regulatory direction of travel supports the category: inland nuclear stations in Russia are required to install cooling towers, with only coastal stations exempted; Uzbekistan has moved toward dry-tower technology after consultation. The principle that inland thermal generation must stop treating rivers as heat sinks is already established in regulatory practice; what has been missing is a rejection path that neither consumes water nor loses capability in the heat.
Reference node used throughout the ARBOK cooling series: 10,000 m³/h of circulation at 8,000 operating hours per year, permitting direct comparison with wet-tower practice.
Wet tower (classical), annual cost: electricity for pumps and fans, $576,000–672,000; makeup water at 3.8% of circulation and $1.50 per m³, approximately $4,560,000 at a delta near 16 °C, rising to approximately $10,440,000 at a delta of 35–37 °C; chemistry, biocides, laboratory work, cleaning and repair, $160,000–480,000. Total: approximately $5,296,000–5,712,000 per year at the lower delta and $11,176,000–11,592,000 per year at the higher one.
Closed vacuum cycle, annual cost: the water line is zero, because water is charged once and circulates without loss. Electricity after the TERU reduction, $230,000–384,000; planned service, $200,000–300,000. Total: approximately $430,000–684,000 per year, and it does not inflate with delta.
Capital cost: a comparable classical node — tower, basin, pumping station and installation — commonly budgeted at up to $10,000,000. Modular container equipment for the same duty: $4,000,000–6,000,000, a capital advantage of $4,000,000–6,000,000 at the outset. Payback is about 7 years — the standard horizon for industrial cooling assets — against operating savings of approximately $4,612,000–5,282,000 per year at the lower delta; at the higher delta, savings exceed $10,000,000 per year.
Avoided-derating value (site-specific, larger than the operating-cost terms above). A 1.3 GW unit halved for one week of heat forgoes over 100,000 MWh — 20–40 million € of foregone revenue in a single week at the 200–400+ €/MWh prices that prevail in those hours, repeated several times per season; an operator who has sold capacity forward must additionally buy replacement energy on the spot market at the same peak prices. Fleet-wide, river-condition curtailment is estimated at approximately 0.5 TWh per year for one national nuclear fleet — roughly 0.3% of its annual generation — worth on the order of 200 million € per year at 200 €/MWh.
Source figures diverge on two parameters. Water-treatment energy intensity is given as 0.5–1 kWh/m³ in the preprint and as 2–3 kWh/t in the companion article; TERU's grid-draw reduction is given as "up to 20%" in the preprint and as "up to a quarter" in the companion materials. Reconciliation against a single engineering figure [требует уточнения из базы].
Mode A is presented as regulator-transparent, not regulator-exempt. No modification occurs inside the licensed process boundary, but a nuclear or comparable safety regulator may still require case-by-case sign-off before an intake-side installation is commissioned, even where no licensed system is touched.
Deployment evidence is proposed, not yet confirmed installed. The source material frames PowerCool as the retrofit path for the 2026 Danube events; it does not report a completed station-scale installation.
Conservative procurement cycles. Power-plant operators, like other heavy-industrial buyers in the ARBOK cooling series, run long decision timelines; the deployment path from proposal to commissioning is not the fast cycle the marketing materials imply.
Site-specific hydraulic integration. Large legacy condenser circuits require site-specific engineering for Mode A tie-in and, more so, for a Mode B conversion.
ARBOK-CoolTower · Arbok-Criojet · TERU · ARBOK-SONAR · ARBOK-ClosedTower
Related technologies
The Optical-Quantum Computer is a next-generation computing architecture based on Digital Light Transistor (DLT) technology — a platform where light controls light with no involvement…
is a compact, autonomous hardware antivirus and network protection device designed to provide physical-layer security against malware, unauthorized access, and DDoS attacks.
is a compact, chip-integrated photonic system designed to perform ultra-fast, analog optical signal processing — including Fourier transforms, spectral decomposition, and temporal filtering — at…

Partnership pathway