BT
B. Trajanoski
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
1 records found
1
Large-Scale Hydrogen Liquefaction: Cryogenic Cooling and Boil-off Gas Recovery
Process Modeling and Techno-Economic Analysis of Catalytic Plate-Fin Heat Exchangers and Ejector-Driven Cycles
Hydrogen is at a turning point in the global energy transition. Its high specific energy makes it attractive for weight- and range-constrained sectors such as long-distance transport and aviation, where liquid hydrogen (LH2) provides the required compact storage medium. Liquefaction, however, remains the main challenge, as industrial plants consume 12-15 kWh per kilogram of liquid hydrogen, roughly three to four times the thermodynamic minimum. Conceptual designs promise to reduce this consumption, but they rest on idealized models. This thesis quantifies two of these idealizations for an 86 tonnes-per-day reference process. The first is the property and kinetic modeling of the cryogenic catalytic plate-fin heat exchanger (PFHX). The second is the assumption of full liquid yield, which ignores the boil-off gas (BOG) generated during storage and truck loading. An ejector-driven recovery cycle is then proposed and assessed for the BOG.
For the cryogenic cooling stage, the helium-neon refrigerant is modeled with an improved equation of state (SAFT-VRQ-Mie) and residual entropy scaling. This replaces the dilute-gas correlations used in earlier studies. The revised thermal conductivity diverges from those correlations by a factor of two at 30 K and underpredicts the measured mixture property by 20-22%. Because the cold-side heat transfer scales as the two-thirds power of conductivity, this increases the required heat exchanger length by about 30% to 7.8 m, and the updated ortho-para conversion kinetics push it to the 8.2 m single-unit manufacturing limit.
For the LH2 storage stage, validated two-phase models show that the loading process dominates the boil-off losses and vents about 2.3% of each delivered load. A one-dimensional ejector model then shows that this BOG, together with the separator flash vapor, can be entrained and returned to the cycle in a single pass. This raises the net liquid product from 86 to 104 tonnes per day. In the adapted cycle, the binding constraint is the downstream PFHX and not the ejector. At the original 75 bar feed pressure, the heat exchanger exceeds the 8.2 m single-unit limit. Only when lowering the feed pressure to 40 bar, which also removes one feed-compression stage, it becomes buildable as a single unit at 8.19 m.
A techno-economic analysis of the isolated cryogenic section quantifies the cost of this recovery. Capital cost rises by 24% and specific energy consumption by 56%. The specific liquefaction cost increases from 0.432 to 0.593 USD per kilogram, an increase of 37.4% that holds across all tested cost assumptions. Ejector-based BOG recovery is therefore technically feasible and increases liquid yield, but it is not economically justified within the cryogenic boundary on its own. ...
For the cryogenic cooling stage, the helium-neon refrigerant is modeled with an improved equation of state (SAFT-VRQ-Mie) and residual entropy scaling. This replaces the dilute-gas correlations used in earlier studies. The revised thermal conductivity diverges from those correlations by a factor of two at 30 K and underpredicts the measured mixture property by 20-22%. Because the cold-side heat transfer scales as the two-thirds power of conductivity, this increases the required heat exchanger length by about 30% to 7.8 m, and the updated ortho-para conversion kinetics push it to the 8.2 m single-unit manufacturing limit.
For the LH2 storage stage, validated two-phase models show that the loading process dominates the boil-off losses and vents about 2.3% of each delivered load. A one-dimensional ejector model then shows that this BOG, together with the separator flash vapor, can be entrained and returned to the cycle in a single pass. This raises the net liquid product from 86 to 104 tonnes per day. In the adapted cycle, the binding constraint is the downstream PFHX and not the ejector. At the original 75 bar feed pressure, the heat exchanger exceeds the 8.2 m single-unit limit. Only when lowering the feed pressure to 40 bar, which also removes one feed-compression stage, it becomes buildable as a single unit at 8.19 m.
A techno-economic analysis of the isolated cryogenic section quantifies the cost of this recovery. Capital cost rises by 24% and specific energy consumption by 56%. The specific liquefaction cost increases from 0.432 to 0.593 USD per kilogram, an increase of 37.4% that holds across all tested cost assumptions. Ejector-based BOG recovery is therefore technically feasible and increases liquid yield, but it is not economically justified within the cryogenic boundary on its own. ...
Hydrogen is at a turning point in the global energy transition. Its high specific energy makes it attractive for weight- and range-constrained sectors such as long-distance transport and aviation, where liquid hydrogen (LH2) provides the required compact storage medium. Liquefaction, however, remains the main challenge, as industrial plants consume 12-15 kWh per kilogram of liquid hydrogen, roughly three to four times the thermodynamic minimum. Conceptual designs promise to reduce this consumption, but they rest on idealized models. This thesis quantifies two of these idealizations for an 86 tonnes-per-day reference process. The first is the property and kinetic modeling of the cryogenic catalytic plate-fin heat exchanger (PFHX). The second is the assumption of full liquid yield, which ignores the boil-off gas (BOG) generated during storage and truck loading. An ejector-driven recovery cycle is then proposed and assessed for the BOG.
For the cryogenic cooling stage, the helium-neon refrigerant is modeled with an improved equation of state (SAFT-VRQ-Mie) and residual entropy scaling. This replaces the dilute-gas correlations used in earlier studies. The revised thermal conductivity diverges from those correlations by a factor of two at 30 K and underpredicts the measured mixture property by 20-22%. Because the cold-side heat transfer scales as the two-thirds power of conductivity, this increases the required heat exchanger length by about 30% to 7.8 m, and the updated ortho-para conversion kinetics push it to the 8.2 m single-unit manufacturing limit.
For the LH2 storage stage, validated two-phase models show that the loading process dominates the boil-off losses and vents about 2.3% of each delivered load. A one-dimensional ejector model then shows that this BOG, together with the separator flash vapor, can be entrained and returned to the cycle in a single pass. This raises the net liquid product from 86 to 104 tonnes per day. In the adapted cycle, the binding constraint is the downstream PFHX and not the ejector. At the original 75 bar feed pressure, the heat exchanger exceeds the 8.2 m single-unit limit. Only when lowering the feed pressure to 40 bar, which also removes one feed-compression stage, it becomes buildable as a single unit at 8.19 m.
A techno-economic analysis of the isolated cryogenic section quantifies the cost of this recovery. Capital cost rises by 24% and specific energy consumption by 56%. The specific liquefaction cost increases from 0.432 to 0.593 USD per kilogram, an increase of 37.4% that holds across all tested cost assumptions. Ejector-based BOG recovery is therefore technically feasible and increases liquid yield, but it is not economically justified within the cryogenic boundary on its own.
For the cryogenic cooling stage, the helium-neon refrigerant is modeled with an improved equation of state (SAFT-VRQ-Mie) and residual entropy scaling. This replaces the dilute-gas correlations used in earlier studies. The revised thermal conductivity diverges from those correlations by a factor of two at 30 K and underpredicts the measured mixture property by 20-22%. Because the cold-side heat transfer scales as the two-thirds power of conductivity, this increases the required heat exchanger length by about 30% to 7.8 m, and the updated ortho-para conversion kinetics push it to the 8.2 m single-unit manufacturing limit.
For the LH2 storage stage, validated two-phase models show that the loading process dominates the boil-off losses and vents about 2.3% of each delivered load. A one-dimensional ejector model then shows that this BOG, together with the separator flash vapor, can be entrained and returned to the cycle in a single pass. This raises the net liquid product from 86 to 104 tonnes per day. In the adapted cycle, the binding constraint is the downstream PFHX and not the ejector. At the original 75 bar feed pressure, the heat exchanger exceeds the 8.2 m single-unit limit. Only when lowering the feed pressure to 40 bar, which also removes one feed-compression stage, it becomes buildable as a single unit at 8.19 m.
A techno-economic analysis of the isolated cryogenic section quantifies the cost of this recovery. Capital cost rises by 24% and specific energy consumption by 56%. The specific liquefaction cost increases from 0.432 to 0.593 USD per kilogram, an increase of 37.4% that holds across all tested cost assumptions. Ejector-based BOG recovery is therefore technically feasible and increases liquid yield, but it is not economically justified within the cryogenic boundary on its own.