Global water scarcity, exacerbated by increasing populations, significantly impacts both domestic and industrial sectors. Industrial water usage, almost double that of domestic use, is projected to increase substantially in coming decades. This increased industrial activity often leads to higher CO₂ emissions, further contributing to climate change and worsening water scarcity. Water and steam are widely used heating mediums in various industries like power plants and chemical plants, but energy efficiency is often compromised due to the release of hot water and steam, particularly steam, which contains substantial latent heat (2.442 MJ kg⁻¹ at 25 °C). Conventional processes release this steam, representing a significant loss of both energy and water. This study focuses on recovering this lost resource. Global natural gas consumption in 2018 produced an estimated 15.7 × 10¹⁸ J yr⁻¹ of latent heat in steam alone, highlighting the potential energy savings through steam recovery. In Japanese incinerator plants, a significant portion (approximately 50 wt%) of domestic waste is water, which converts to steam during combustion at 900 °C. Additional steam is produced during waste combustion (around 30 wt%), and to prevent dioxin formation, large quantities of water (nearly double the waste weight) are sprayed for cooling, further increasing steam release. This leads to visible steam plumes, which are a public concern hindering the construction of incinerator plants in urban areas. This paper proposes a system to address these issues by incorporating a steam recovery membrane unit, enabling water and energy self-sufficiency in waste incinerator plants.

Existing literature highlights the challenges of high-temperature steam recovery. Conventional organic membranes (polyamide, polyimide, polyvinyl alcohol) lack the thermal stability needed for temperatures exceeding 150 °C. Inorganic membranes, such as LTA zeolite and amorphous silica, exhibit instability under hydrothermal and/or acidic conditions. While recent advancements in membranes derived from SOD, FAU, MOR, MFI zeolite, metal-doped silica, and some perfluoro-polymers show selective steam permeation at high temperatures, their upper boundaries of steam permeance remain limited (around 10⁻⁶ mol m⁻² s⁻¹ Pa⁻¹). Previous research by the authors demonstrated that organosilica membranes with organically linked silsesquioxane structures exhibit significantly higher steam permeance (2-5 × 10⁻⁶ mol m⁻² s⁻¹ Pa⁻¹) and high steam/nitrogen permeance ratios (several hundred to several thousand) at 150-200 °C, offering enhanced hydrothermal stability. This superior performance makes them suitable candidates for the steam recovery system.

This study involved membrane fabrication, laboratory-scale long-term stability testing, bench-scale steam recovery testing in a running incinerator plant, and process simulation. Organosilica membranes (M-1 and M-2) were fabricated via the sol-gel method using bis(triethoxysilyl)ethane (BTESE). The membranes consisted of a layered structure: a support layer (α-alumina porous tube), an intermediate layer (BTESE-derived organosilica), and a separation layer (BTESE-derived organosilica) with a pore size of about 1 nm. Laboratory-scale tests evaluated single-gas permeance and long-term stability under simulated waste streams (varying compositions and temperatures) for up to 190 days. The effect of HCl concentration on membrane performance was also assessed. Bench-scale tests involved a steam recovery demonstration in an operational waste incinerator plant for two days, measuring temperature, pressure, steam molar fraction, recovered water flux, and steam permeance. Process simulation, using counter-current mode, was conducted using waste stream properties from the plant and membrane performance data from laboratory tests to assess the system’s effectiveness in terms of steam recovery, effective enthalpy recovery, dimensionless steam recovery, and retentate dew point. Membrane characterization included single component permeation (steam, He, H₂, CO₂, N₂, CH₄, CF₄, SF₆) to determine permeance using a film-flow meter for gases and a cold trap for steam. For multi-component permeation, an in-house apparatus allowed for the evaluation of steam permeance from simulated waste gas streams (H₂O, N₂, O₂, HCl). The permeance was calculated using formulas that account for partial pressures, the permeating flow rate, and the membrane area. The simulation model assumed a plug flow and used mass balance equations for each component (H₂O, N₂, O₂, CO₂) in both upstream and downstream to evaluate component flow rates based on permeance values. The effective enthalpy recovery was calculated considering latent heat recovery, sensible heat recovery, and isothermal compression energy. The dew point of the retentate was calculated using the Antoine equation.

Laboratory-scale testing demonstrated the excellent molecular sieving properties of the organosilica membranes, with permeance decreasing dramatically with increasing molecular size. The membranes exhibited high hydrothermal, oxidative, and acidic stability, maintaining stable performance for 190 days under varying conditions, including exposure to 400 ppm HCl. Steam permeance remained consistently high, significantly better than other reported membranes. Bench-scale tests in the running incinerator plant successfully demonstrated steady steam recovery over two days, confirming the laboratory findings' applicability to real-world conditions. The steam permeance values obtained in the plant (several 10⁻⁶ mol m⁻² s⁻¹ Pa⁻¹) were consistent with those obtained in the laboratory. The process simulation predicted that the system could recover up to 74 t/day of steam with a membrane area of approximately 1500 m², resulting in significant energy recovery (approximately 70% of waste combustion energy) and achieving water self-reliance for the incinerator plant. The dew point of the retentate remained significantly lower, eliminating the visible steam plume.

The findings demonstrate the feasibility and effectiveness of the proposed steam recovery system using organosilica membranes. The exceptional hydrothermal, oxidative, and acidic stability of the membranes, coupled with their high steam permselectivity, addresses the limitations of existing membrane technologies for high-temperature applications. The successful bench-scale demonstration validates the system's performance in a real-world setting. The process simulation further reinforces the system's potential for significant energy and water savings in waste incinerator plants. This approach offers environmental benefits (reduced water consumption and greenhouse gas emissions) and economic benefits (energy cost reduction). The elimination of the steam plume addresses public concerns, facilitating the construction of incinerator plants in densely populated areas. The system's applicability extends beyond waste incineration plants to other industrial sectors with high-temperature steam streams.

This study successfully demonstrated a novel steam recovery system using hydrothermally stable organosilica membranes, achieving both energy and water recovery. The system’s effectiveness was validated through rigorous laboratory and bench-scale testing and process simulation. Future research could explore optimizing membrane design and exploring different configurations to enhance performance further and investigate the scalability and economic viability of this technology for broader industrial applications.

The current study focused on a specific type of waste incinerator plant in Japan. While the findings are promising, further research is needed to assess the system's applicability to different waste compositions and incinerator plant designs. The process simulation made certain simplifications, such as neglecting low-concentration contaminants, which could be considered in future, more detailed models. Long-term operational data beyond six months of operation would further strengthen the conclusions regarding the membrane's long-term durability and performance.