What if the coolant that gives a metal hydride reactor its shortest hydrogen-storage cycle also demands nearly six times the pumping power? This study finds that pure ethylene glycol can cut predicted cycle time, but the trade-off is substantial.

The study numerically investigates a hydrogen storage metal hydride reactor using ethylene glycol-water mixtures as heat transfer fluid. Its main outcome is hydrogen storage cycle time, which reflects absorption and desorption kinetics in the metal hydride reactor. Hydrogen is recognized as a promising clean energy carrier, with demand expected to increase by around 30 percent between 2015 and 2023 because of industrial and transport applications.

Batteries are increasingly used for short-distance or light transport, while hydrogen is preferred for long-distance, heavy transportation and industrial uses because of higher gravimetric energy density and refueling time. Safe storage with easy reversibility remains a major challenge, and solid-state storage based on metal hydrides has attracted attention because of safety, reversibility, and compactness.

The performance of MmNi4.6Al0.4 is frequently hindered by sluggish heat transfer during hydrogenation and dehydrogenation. Solid-state storage can provide high volumetric density and reversible release, but metal hydride materials have naturally low thermal conductivity that hampers heat transfer during hydrogen uptake and release.

Ethylene glycol-water combinations allow thermophysical properties to be adjusted by changing the mixture proportion, making them candidates for cooling and heating purposes in industry. There is a lack of studies on how EG-water mixture composition affects the kinetics and thermal behavior of metal hydride hydrogen storage reactors.

Thermal management is difficult because hydrogen absorption is highly exothermic, whereas desorption is strongly endothermic. The study examines ethylene glycol and its aqueous solutions, with EG-water coolant compositions varying from zero to one hundred percent EG concentration.

Ethylene glycol is widely used because of its high boiling point, low freezing point, and chemical stability suitable for low-temperature applications. Because ethylene glycol is completely miscible with water, blending can tailor thermal performance.

Ethylene glycol has low thermal conductivity, about zero point two five watts per meter-kelvin, and high dynamic viscosity, so larger flow rates and pumping power are required. The physical model is one tube with multiple layers, with a hydride bed maximum diameter of 35 millimeters and length of 450 millimeters.

A 12-millimeter diameter region is empty for hydrogen gas, while a surrounding layer contains metal hydride powder. The outermost shell is about 8 millimeters thick and accommodates the heat transfer fluid; the reactor contains 1.8 kilograms of MmNi4.6Al0.4 alloy.

The longitudinal view shows arrows carrying cooling fluid along both the upper and lower outer channels, while the porous filter separates those channels from the metal-hydride regions. In the cross-section, the cooling fluid forms a continuous annulus around the hydride bed, with the central opening reserved for hydrogen gas.

Hydrogen absorption in the metal hydride reactor triggers an exothermic reaction that generates heat within the system. The cooling fluid therefore plays a crucial role in managing heat generated during hydrogen absorption. The rate at which cooling fluids absorb heat and the amount of heat they dissipate are directly proportional.

The analysis is simplified using several assumptions. The analysis begins by introducing several assumptions, specifically to simplify the analysis and make the reasoning that follows more manageable. The model includes an equation for the combined energy of the hydrogen gas and the hydride bed.

That combined-energy equation represents the energy treatment used for the hydrogen gas and hydride bed. During absorption, heat transfers from the wall of the bed to the external cooling fluid. For an elemental area of the reactor wall in an axial position, the heat transfer rate between the reactor wall and external fluid is given by the model.

Viscosity, thermal conductivity, density, and specific heat capacity vary with heat transfer fluid temperature, so temperature-dependent fluid properties were included in the numerical model. For ethylene glycol and its mixtures, viscosity can vary by one order of magnitude or more between 20 degrees Celsius and 100 degrees Celsius.

The heat transfer coefficient at the coolant-reactor interface was held constant at 1000 watts per square meter-kelvin for every coolant composition. This assumption investigates different EG-water thermophysical properties under the same heat transfer conditions, so differences in absorption, desorption, and cycle times result from different coolant characteristics.

The hydride-bed governing equations were discretized using the Finite Difference method, and a transient stability criterion for two-dimensional heat conduction determined the simulation time step. The numerical simulations were performed in MATLAB, and predictive capability was evaluated by comparing simulation results with published experimental data.

The model’s R2 value was 0.987 based on simulated and experimental average bed temperatures for hydrogen absorption, indicating high correlation and excellent agreement. A grid independence test evaluated how spatial discretization affected predicted bed temperature by comparing a 41 by 1001 grid with a finer 61 by 1251 grid.

The maximum bed-temperature difference was zero point 13 percent, so the 41 by 1001 grid was used for all later simulations. The numerical model was validated against experiments for hydrogen absorption and average metal hydride bed temperatures, with good agreement under similar operating conditions despite some initial deviations.

The results used fixed hydrogen supply and discharge pressures of 30 bar, with heat transfer fluid inlet temperatures of 298 kelvin for absorption and 380 kelvin for desorption. The literature experiments used a water-cooled setup, after which the same model was adapted for EG-water mixtures using temperature-dependent thermophysical properties.

The current study is purely numerical, with no independent experiments performed for its purposes. In the main plot, the open experimental circles generally follow the solid model curve: both rise sharply at the beginning and then decline toward the initial-temperature level.

The largest visible separation occurs during the early and middle portions of the cooldown, while the points cluster closely with the curve after roughly two hundred seconds. Validation was performed only for water-cooled hydrogen absorption because experimental data for desorption and EG-water cooling mixtures could not be found in the literature.

The validation therefore confirms prediction of transient thermal behavior only under the validated water-cooled absorption conditions. Results for desorption and other cooling-fluid mixtures are purely numerical predictions. The study examined cycle time, absorption time, and desorption time as supply pressure varied from 20 to 40 bar.

Increasing supply pressure increased cycle time and desorption time, while absorption time did not significantly decrease with increasing weight and supply pressure. A supply pressure of 30 bar was chosen for further study. Looking across the plotted markers, the absorption-time curve declines steadily, while the cycle- and desorption-time curves rise overall, with the cycle-time line showing a slight dip at the second pressure point.

At the far-right point, the cycle-time marker is near two thousand seconds, whereas the desorption-time marker is around one thousand seven hundred seconds and the absorption-time marker is near three hundred seconds. As hydrogen is absorbed and the reaction slows, heat production decreases and heat transfer causes the bed to cool.

Hydrogen uptake continues after the maximum temperature is reached, but the reaction slows as the hydride bed approaches saturation. When heat removal becomes greater than heat production, the bed temperature decreases. The lowest peak bed temperature was associated with superior thermal conductivity and heat removal capability.

Pure ethylene glycol showed a slight increase in peak temperature because of its low thermal conductivity. Peak bed temperature alone is insufficient to assess reactor performance. In the enlarged cooling region, the curves remain tightly grouped through the temperature drop, but their endpoints are visibly staggered in time.

The dashed water curve extends farthest to the right, while the EG blends and pure EG reach the baseline earlier, with the separation spanning several tens of seconds. A lower peak temperature does not necessarily correspond to the shortest hydrogen storage cycle.

The complete transient thermal response, including heat removal during absorption and heat supply during desorption, governs overall reactor performance. Peak bed temperature should therefore be interpreted together with absorption time, desorption time, and total cycle time.

Figure 7 assesses heat transfer fluid performance at a target hydrogen concentration of 99 percent, where faster concentration times indicate better heat transfer fluid efficiency. Under the adopted numerical assumptions, pure EG produced the shortest predicted cycle time, a 10.7 percent reduction compared with water.

The 20, 40, 60, and 80 percent EG blends showed lower cycle times with reductions of 3 to 4 percent. In the enlarged early-time region, around two hundred eighty-two to two hundred ninety-four seconds, the curves are visibly separated: the solid black pure-EG line is highest, while the dashed water line is lowest, with the blended formulations between them.

By roughly two hundred ninety-four seconds, the plotted concentrations span approximately zero point nine eight zero to zero point nine eight five. Pure ethylene glycol had the lowest normalized cycle time, zero point 89, indicating improved absorption and desorption characteristics.

That improvement was accompanied by pumping power rising by nearly sixfold because EG has significantly higher viscosity than water. The thermal conductivity of 100 percent EG was only 42 percent of water’s, so its natural heat transfer ability was not as good unless faster flows were used.

Looking across the table, the cycle-time column stays close to the water reference for the intermediate blends, ranging from approximately zero point ninety-five to zero point ninety-seven, while the pumping-power column expands from about two point five to six point zero.

The thermal-conductivity column simultaneously declines across the EG rows, from approximately zero point eighty-nine to zero point forty-two. Transient mathematical modeling studied heat and mass transfer during hydrogen storage in a tubular reactor using MmNi4.6Al0.4, and the model was verified against available experimental data before analyzing EG and EG-water solutions.

An increase in hydrogen supply pressure decreased absorption time because of a stronger driving force, but total cycle duration increased because desorption and thermal recovery took longer. Among the considered coolant compositions, pure EG had the smallest hydrogen absorption, desorption, and cycle times; compared with water, its hydrogen storage cycle time was 1165 seconds, about 10.7 percent lower.

Coolant selection should also consider pumping power, pressure drop, viscosity, economic cost, operating temperature requirements, and the adopted constant external heat-transfer coefficient. Experiments with EG-water mixtures should be performed to confirm the analysis, and the required pumping power should be analyzed.

Pure ethylene glycol produced the shortest predicted hydrogen-storage cycle, 1165 seconds and about 10.7 percent below water, but coolant selection must also account for viscosity, pumping power, pressure drop, cost, and the study’s numerical assumptions.