|
9.3 Conceptual Design of Hydrogen Liquefaction Facilities 9.3.1 R&D Goals In the WE-NET 2nd phase starting this year, development of the large-scale hydrogen liquifier is positioned as a part of long-term development project. The final target of it will be collecting basic data of aerodynamic design and seal design, which are required to develop high-performance hydrogen compressor. We conducted researches on challenges of aerodynamic design and seal design and formulated development plan. With regard to the aerodynamic design technology, aerodynamic performance of impeller and diffuser were studied with the use of fluid analysis, and scale effect was examined quantitatively based on the results of the 1st phase. In connection with the sealing technology to prevent leakage of hydrogen gas, sealing system was studied and selected, and the amount of leakage was roughly estimated. In addition, as short-term development assignments of hydrogen utilization technology, hydrogen car system and hydrogen refueling station were selected. Based on the selection, feasibility study of small-amount hydrogen liquifier was conducted. 9.3.2 Results in FY 1999 9.3.2.1 Study on Aerodynamic Performance of Hydrogen Compressor (1) Aerodynamic performance of impeller In the 1st phase, as a part of preliminary study of hydrogen compressor, which is critical to hydrogen liquefaction basic unit, prototype test of small backward impeller was conducted, and internal flowing was measured with laser velocity meter. With the 25 backward impeller, the efficiency was very high (about 88-90%), which verified that the result of flowing analysis qualitatively agrees with the measuring results. With the use of fluid analysis, studies were conducted to determine the backward angle that is suitable for the hydrogen compressor. 2 types of impeller forms (20, 15) were designed and made three dimensional viscous flow analysis of them. Their flow conditions were compared to one another with 25 and 5 backward impellers, whose performance test had been conducted in the 1st phase. The computational condition for flow analysis was pneumatic test condition (7,800rpm, 1.45m3/s), which was the same as the 1st phase. Figure 9.3.2-2 shows relative flow distribution around exit (indicated section). According to the distribution, low flow area developed on the shroud side of the suction surface. Compared to 25 backward, low flow areas of 20 backward and 15 backward are larger. However, they were smaller than that of 5 backward impeller. As the backward angle becomes smaller, the flow velocity distortion tends to grow bigger. However, the distortion disparity between the 20 backward and 25 backward is small, and high efficiency was achieved by 25. Judging from the results, backward 20 may achieve high pressure ratio at the same efficiency with the 25. Figure 9.3.2-3 shows work input coefficients of the impeller determined by flow analysis. Compared to 25 backward impeller, coefficients of backward impellers with the angles of 20, 15 and 5 are bigger by 3.4%, 7.1% and 16.5% respectively. Provided that 20 backward impeller could achieve the same efficiency as 25backward impeller, pressure ratio would rise about 0.5% due to an increase in work input coefficient, which leads to reduction of stages by 3%. If the pressure ratio of a single stage is unified, the diameter of the impeller can be reduced by about 2%. (2) Aerodynamic Performance of Diffuser To determine the diffuser that is suitable for hydrogen compressor, performance study was conducted with the use of two-dimensional viscous flow analysis. In the 1st phase, we tried to understand interflow of double circular arc profile diffuser and NACA65 type profile diffuser by flow analysis. Figure 9.3.2-4 shows flow distribution around the maximum efficiency point, results of the test in the 1st phase. In case of the double circular arc profile diffuser, low flow area developed from flow direction 2/3 on the suction surface. As a result, the width of flow passage was smaller than the geometrical flow passage width and the flow did not slow down in the second half of the flow passage. On the other hand, NACA65 type profile diffuser has slow flow area at the exit side of the acting face, but the flow will slow down toward the exit of the wing. If the wing inlet angle is made wider to increase flow, low flow area will be smaller, which may unify the flow and increase efficiency. (3) Scale Effect Along with the make up compressor and the recycle compressor, the hydrogen compressor is high-pressure multistage type. From the 1st stage to the last stage, the impeller diameter varies between 1000mm and 300mm. Scale effect on the performance was studied. With regard to the scale effect, there are many reports of Reynolds number effect. In this project, efficiency difference was determined by using appraising formula, which was proposed by Casy.1) In this method, inside turbulence flow, which equivalents to impeller, was defined and its friction coefficient was calculated to take Reynolds number effect into account. The result is shown in Figure 9.3.2-5. Provided that exit width ratio is b2/D2=0.08 (b2: impeller exit width, D2: impeller diameter), the efficiency difference will be about 2.6% between impeller diameter of 1,000mm and 300mm. 9.3.2.2 Studies on Hydrogen Gas Sealing (1) Studies on sealing form and sealing method Since the hydrogen is the flammable gas whose molecular weight is light and it is easy to leak. In terms of safety and hydrogen liquefaction basic unit increase, sealing technology is important. To select the optimum sealing method, researches were conducted on 3 methods; 1) spiral groove sealing, 2) magnetic fluid sealing; and 3) wet sealing. Spiral groove sealing, outlined in Figure 9.3.2-6, was selected in the end. Suitability of each method was tested on the assumption of the high-pressure recycling compressor, which has long impeller diameter at the first stage, and has the largest difference between internal and external pressure of the compressor at the last stage. As a result, magnetic fluid sealing did not meet the standards of allowable pressure difference and peripheral velocity. Wet sealing did not meet the standard of allowable peripheral velocity. On the other hand, spiral groove sealing met the both standards of allowable pressure difference peripheral velocity. Spiral groove sealing is a kind of noncontact sealing. Between rotating ring, with helical ditch, fixed to the shaft and fixed ring fixed to the casing, a few-micron clearance is formed due to dynamic pressure caused by rotation. Leak gas supplies buffer gas and is collected with buffer gas. (2) Studies on Hydrogen Gas Leak As to the selected spiral groove sealing, hydrogen leakage was roughly estimated. With the last stage of high-pressure recycling compressor (the difference between internal and external pressure of the compressor: 40MPa) assumed, the leakage was calculated. The results are shown in Figure 9.3.2-7. Provided that the clearance gap had been 4mm, functional clearance without contacts of sealing, the leakage was 82.5Nl/min. Similarly, leakage of the first stage was 3.7 Nl/min. Total leakage of the all stages was estimated according to the mean value, which was determined by averaging the values of the last stage and the first stage. The result was about 0.009% of the full flow for the high-pressure recycling compressor. Judging from this, leakage is very small compared to the full flow, and spiral groove sealing is effective. 9.3.2.3 Studies on Small Capacity Hydrogen Liquefier (1) Hydrogen liquefaction cycle Assuming that the scale of hydrogen liquifier that supplies liquid hydrogen to hydrogen car stations was (10-30) t/d, studies were conducted on the same process cycle as 11t/d liquifier, which had been manufactured in the past. The process flow is outlined in Figure 9.3.2-8. As an expansion turbine used for the liquefaction process, the largest one of existing turbines can be used. However, the existing expansion turbine does not have power recovery system, therefore, it did not recovery power. (2) Hydrogen Compressor With regard to the expansion turbine that is used in the liquefaction process, a reciprocating compressor and an oil cooling screw compressor were used to study with 30t/d. It is possible to produce the reciprocating compressor with the capacity of 30t/d. If the capacity is more than that, the initial cost will rise in accordance with the increase in capacity. On the other hand, the oil cooling screw compressor with the capacity of 30t/d can be also produced. However, the power rose because it compressed oil, too. (3) Nitrogen Reliquefaction In case of 30t/d nitrogen reliquefyier, the same process as 300t/d liquifyier was taken as is shown in Figure 9.3.2-9. For liquefaction capacity of 10t/d, the process using one unit of expansion turbine was adopted to study on required power and so on. (4) Required Power and Basic Unit of Power With liquefaction capacity of 10t/d and 30t/d, estimated results of power basic unit and others are in Table 9.3.2-1. In case of 30t/d, total power, process efficiency and power basic unit were 13.28MW, 37.3% and 0.955kW/Nm3 respectively. Incidentally, results of the first period were 108.5MW, 45.6% and 0.779kW/Nm3 respectively. In order to reduce power basic unit of (10-30) t/d further, power reduction of the compressor and power recovery of the expansion turbine are demanded. 9.3.3 Research Plan for FY 2000 Development of the large-scale hydrogen liquefyier is required to continue as a long-term research project. From now on, the focus of the research should be shifted to aerodynamic performance for developing the large-scale hydrogen compressor and collection of basic design data associated with sealing performance. With regard to aerodynamic performance, it is required to conduct tests for verifying optimum form of the impeller. It is also necessary to conduct prototype tests on impellers with various diameters for verifying the size effect. As for sealing, it is required to verify performance and durability of the selected spiral groove sealing through a basic test using real hydrogen gas. As a short-term research project, small-capacity hydrogen liquifyier was studied this year. It is necessary to work on this project emphatically, as the hydrogen car industry demands it strongly. In this context, we would like to conduct the cost estimation and the research about selection of hydrogen compressors for relatively small hydrogen liquifyier system to come up with technical challenges and development objectives. Reference 1) Casy M. V.; "The Effects of Reynolds Number on the Efficiency of Centrifugal Compressor Stages", ASME Paper 84-GT-247, 1984.
  |
