While the invention is susceptible to various modifications and alternative forms, specific embodiments are shown by way of example in the drawings and are described in detail herein. It should be understood, however, that the invention is not intended to be limited to the particular forms disclosed. Rather, the invention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention as described.
Referring to
In one embodiment, the cooling fluid will pass through each heat sink exchanger to a separate water-to-air heat exchanger, where the water is cooled and returned to the liquefier system to cycle through the heat sink exchangers again. In another embodiment, a common heat sink exchanger may be used that is capable of handling the heat rejected from each AGRR stage and connected to a closed loop water chiller. This water chiller will have a circulation pump for the water as well as a fan to drive air convection through the water radiator. The pump and fan powers for the chiller will be small and supplied by small motors. The process stream heat exchangers (15) and the cryogenic AGRR cold heat exchangers (14) may be contained within a cold box or otherwise thermally isolated from the surroundings to reduce parasitic heat leaks into the cryogenic liquid product (18).
The designed cooling capacity of the AGRR stages scales with the rate of production of LNG or LH2. The AGRL design can be scaled from several hundred or thousand gallons/day upwards to much larger liquefaction capacities. This type of AGRL for natural gas or for hydrogen has been designed to have a FOM within the range of about 0.52 to 0.69.
Referring to
After temperature gradients are established in arrays of tubes (21) during a short startup sequence, the AGRR stage (12) operates as follows: A piston (28) expands the refrigerant (23) in all the small tubes (21) of one of the dual regenerators (the right regenerator in
In the embodiment depicted in
During the cycle, the average temperature of the refrigerant (23) in the array of tubes (21) is increased in one regenerator (20) and decreased in the other regenerator (20) as the heat transfer fluid (35) flows. The direction of flow of the heat transfer fluid through the active regenerator (20) is reversed by the displacer or circulator (32) when the temperature of the bottommost layer of tubes (21) increases by about half the temperature decrease of the small tubes caused by the polytropic expansion of the refrigerant in bottommost layer of tubes in the active regenerator (20). This flow reversal of the heat transfer fluid is synchronized with the compression/expansion of the refrigerant in the dual regenerators.
The work required to pump the heat from the thermal load to the heat sink is distributed over all the tubes (21) comprising the dual active regenerators (20) of each AGRR stage. By coupling the pistons (28) together via a direct linkage, the net work of refrigerant compression that must be externally supplied by the drive mechanism (30) is the net work required for the thermodynamic refrigeration provided by the AGRR stage. The offset of most of the work of compression of the refrigerant by work recovery from simultaneous expansion of the refrigerant and the distributed work input as a function of temperature are two of the fundamental reasons for a high efficiency in an individual AGRR stage (12) of an AGRL. The cooling power of the AGRR stage (12) is proportional to the heat transfer fluid flow rate and the effective temperature changes of the refrigerant caused by compression/expansion. As described, the heat transfer fluid (35) is cooled or heated by the effective temperature changes fluid, helium gas at a modest pressure, (e.g., 1-2 MPa), in the process stream heat exchanger for each AGRR stage. The thermal cooling load at each stage is related to the He and H2 mass flow rates via the equation
{dot over (m)}
He
c
He
ΔT
cold
={dot over (Q)}
load
={dot over (m)}
H2(hf−hi)
where cHe is the heat transfer fluid heat capacity, hf and hi are enthalpies of the H2 at the entrance and exit of the process stream heat exchanger for the respective AGRR stage, ΔTcold is the effective temperature change of the He heat transfer fluid as it passes in counterflow through the process stream heat exchanger with an average H2 exit temperature of about Tcold for that particular process stream heat exchanger. After passing through the process stream heat exchanger, the warmer He heat transfer fluid then flows back through the other dual active regenerator of the AGRR stage where the refrigerant within all tubes in this array is compressed with corresponding effective adiabatic temperature increases above the mean temperature at a particular longitudinal position along the active regenerator. At the beginning of this “hot blow” period, the temperature spanned by the dual active regenerator with the compressed refrigerant in each tube is ˜Tcold to ˜Thot+ΔThot. The heat transfer fluid picks up heat from each of the tubes in the active regenerator as it flows and eventually leaves the hot end of the regenerator at a temperature higher than the mean heat sink temperature.
The periodic motion of the heat transfer fluid is synchronous with the operation of the refrigerant-filled tubes and only shifted in phase by the ratio of the time for the compression/expansion step to the time for the blow periods (this ratio is usually small, i.e., 0.05 to 0.1). To accomplish this effectively with a single room temperature heat transfer fluid displacer or circulator, the preferred embodiment uses a valve arrangement to create a periodic of the refrigerant (23) in the array of tubes (21) of the active regenerators (20) caused by the compressor/expander assembly (24).
The frequency of periodic flow reversal in the displacer or circulator (32) and the operation of the three-way valves (36) is properly phased with the temperature changes in the array of tubes (21) and can typically operate at reasonable frequencies near 1 Hz. The dual regenerator configuration in this AGRR stage (12) allows the heat transfer fluid loop to be hermetic and reversible according to the motion of the displacer or circulator (32) and pistons (28). The dual regenerators (20) are identical and operate 180° out of phase with each other. In other words, the compression of the refrigerant (23) by a piston (28) connected via manifold (26) to the array of tubes (21) in one of the dual regenerators (20) is synchronous with the expansion of the refrigerant (23) by a piston (28) connected via a manifold (26) to the other dual regenerator (20). Similarly, the heat transfer fluid flow or cold blow (from top to bottom) in one active regenerator (20) is synchronous with the heat transfer fluid flow or hot blow (from bottom to top) in the other regenerator (20). The heat transfer fluid (35) reciprocatively flows through each of the dual regenerators (20) while flowing semi-continuously in counterflow through the cold heat exchanger (14) coupled to the process stream heat exchanger (15) and the heat rejection exchanger (13) coupled to the heat sink exchanger (17).
In one embodiment, each passive micro-regenerator (22) for each individual tube in a given layer of tubes in the two-dimensional array of tubes (21) comprising the dual active regenerators (20) has a thermal mass of ˜30-50 times the thermal mass of the refrigerant (helium gas) that flows in and out of each tube (21) during compression or expansion. A typical passive micro-regenerator material is stainless steel spheres with a diameter of ˜200-300 microns (0.2-0.3 mm). The temperature span across each layer of passive micro-regenerators (22) feeding each layer of tubes (21) in the active regenerator will be from the average cold temperature of the refrigerant plus tube combination in that layer of tubes (21) and the near room temperature refrigerant in the manifold (26) connecting the compressor/expander to the active regenerator. The diameter of each cylindrical passive micro-regenerator (22) is the same as the diameter of the small tube it is connected to in the dual active regenerators. The length of these passive micro-regenerators (22) is a design variable and was chosen as a few centimeters in one tested embodiment. The pressure drop for the helium gas refrigerant as it flows through the passive micro-regenerators (22) in and out of the individual tubes (21) was designed to be very low at an operational cycle time of ˜1 Hz. The pressure changes of the helium refrigerant (23) within the tubes (21) can be from ˜215 psia to ˜430 psia in a typical operation. The mean pressure of the He heat transfer fluid that flows through the dual active regenerators (20) and the heat transfer fluid in the reversible displacer or circulator (32) is an operating variable and was about 200 psia in a test embodiment. It may be higher or lower if desired. The pressure drop of the heat transfer fluid (helium gas in the preferred embodiment for LH2) through optimally designed dual regenerators (20) is typically very small, i.e., 10's to 100's of Pa.
The pressure versus temperature diagram of
In the hot blow, the heat transfer gas comes out of the right end of the active regenerator at a temperature Thot+φΔThot where φ ranges from 1 to 0 during the blow period (usually φ averages about 0.5) and ΔThot is the temperature change of the hottest layer of tubes. This hot gas can reject heat to the cooling fluid as the heat transfer fluid cools back toward Thot. Similarly, during the cold blow, the heat transfer gas comes out of the left end of active regenerator at a temperature of Tcold−φΔTcold and it can absorb heat from the thermal load from the process stream as it warms toward Tcold. The operation of this active regenerator is similar to that of high performance regenerators in other regenerative refrigerators with the added feature that each refrigerant-filled tube in the regenerator has the ability to actively change its temperature and thus enable distributed refrigeration within the regenerator rather than just be a passive heat sink/source as is traditionally the case. Hence, there is a large surface area for heat transfer between the tubes and the heat transfer fluid in the preferred embodiment. Also, the pressure drop to flow, the longitudinal (axial) conduction, and porosity of the regenerator are as small as possible in the preferred embodiment. However, other active regenerator configurations may be used in other embodiments. The temperature increases of the refrigerant-filled tubes along the regenerator longitudinal axis must satisfy the second law of thermodynamics and be in the ideal ratio of absolute temperatures along the regenerator as has been schematically illustrated in
The cooling power of the AGRR stage at the cold temperature, the hot sink temperature, the thermal load temperature, the heat transfer fluid pressure, the effective tube wall temperature change of the cold heat transfer fluid leaving the cold end of the dual active regenerators, and the φ factor that specifies the fraction of the effective tube wall temperature change that can be used for a given heat transfer fluid flow period are all design variables that can be set. The flow rate of the He heat transfer fluid, {dot over (m)}He, is calculated from the equation:
{dot over (Q)}C={dot over (m)}HecpΔTCφ
where {dot over (Q)}C is the cooling power at the cold temperature, TC, cp is the heat capacity of He at constant pressure, and ΔTC is the effective tube wall temperature change at the coldest row of refrigerant-filled tubes in AGRR active regenerator. The typical variables for an active regenerator with 50 W of cooling power at 240 K are presented in Table 1 below.
In the preferred embodiment, a rectangular tube array is used with a staggered tube arrangement in successive layers from the cold end to the hot end of the dual active regenerators. The length in the x direction is the active length of the refrigerant-filled tubes, or the longitudinal axis of the regenerator. The length in the y direction is the length of each row of refrigerant-filled tubes where the number of tubes depends upon the tube diameter and separation between each tube in a row. The z direction is the heat transfer fluid flow direction and this length is determined by the tube diameter and the separation between the layers of rows of tubes. The total number of tubes in the dual active regenerators on each side of the AGRR stage is the product of the number of tubes in each row and the number of layers of rows. In one embodiment, the length of x and y have the same value and the z length has a value three times the y length, although other dimensions are possible. The primary independent variable becomes the tube diameter once the rectangular dimensions of the dual regenerators are chosen with the constraints above.
The average He heat transfer gas properties at 200 psia can be calculated at the average temperature of the dual active regenerators. The density, heat capacity, viscosity, and thermal conductivity can be obtained and used to calculate the Prandtl number. The Reynolds number of the heat transfer fluid flow can be determined using the accepted equation from the literature for this geometry. For the various configurations, the heat transfer coefficient, the friction factor, and the effective thermal conductivity can be calculated. These values can then be used to calculate the entropy generated from the three mechanisms described above and the total entropy generated can then be used to obtain the FOM of the AGRL.
Of immediate note is that the AGRR stage efficiencies ranges from a relatively small value in the case with 10 total tubes of 0.635 cm (0.25″) outer diameter, 2.5 cm length, and a reasonable ΔT of 10 K to a very impressive value in the case where 0.15875 cm ( 1/16″) outer diameter, 7.5 cm length, and a higher but achievable ΔT of 15 K. These dimensions along with Table 2 are provided as examples of one embodiment only and should not be construed to limit the scope of the disclosed invention.
In the preferred embodiment, each of the dual regenerators in each AGRR stage has the following characteristics:
high specific area (˜10,000 m2/m3)
very thin wall tubes
appropriate ΔT vs. T characteristics
mechanically strong enough for modest pressures
leak tight and able to withstand cyclic mechanical pressure loads
low pressure drop
high transverse and low longitudinal (axial) conductivity and
mass producible at low/modest cost.
As mentioned, one variable that can be manipulated to affect the efficiency of the AGRR stage is the effective adiabatic temperature change of the refrigerant-filled small diameter thin-walled tubes. For example, a 5/32″ outer diameter×0.003″ wall stainless tubing with a compression pressure ratio of 3 has a possible ΔT of about 19 K with a mean operating temperature of 300 K and an initial He refrigerant pressure of 215 psia.
In the preferred embodiment of a hydrogen AGRL, the thermal load, {dot over (Q)}load, from the hydrogen process stream is transferred via convective heat transfer to the heat transfer flow stream for dual active regenerators from a continuous stream in the circulator and heat rejection exchanger. The pressure drop of the active regenerator and heat exchanger combinations can be kept reasonably low by design choices, so this circulator needs high volumetric efficiency at relatively low head. The extra work for this component of each AGRR stage is modest so its effect on overall efficiency of each stage and FOM of the AGRL is relatively small. The operation of the active regenerator in this AGRR stage is similar to that in any regenerative refrigerator with the added feature that each refrigerant-filled tube in the active regenerator has the ability to provide refrigeration rather than acting only as a passive heat sink/source. The selection of many small diameter tubes in the preferred embodiment creates a large surface area for heat transfer between the refrigerant-filled tubes and the heat transfer fluid, thereby reducing the dominant irreversible entropy mechanism in the active regenerator, in each AGRR stage, and in the AGRL as a whole.
The preferred embodiment of a hydrogen AGRL includes the following features:
A hydrogen AGRL also preferably has the characteristics described in Table 3 for each AGRR stage:
With these total regenerator tube lengths, the geometries of each AGRR stage can be obtained with choices of 1x, 1y, and 1z, the respective lengths of the rectangular regenerator as described above. Table 4 summarizes the selected geometries of each AGGR stage of one embodiment of the AGRL with tubes of diameter 0.3175 cm. Again, these geometries could be varied.
For purposes of this description, the refrigerant (23) in the upper dual active regenerator (20) of
Each of the three-way valves (36) can be switched, and the pistons (28) of the compressor/expander expand the refrigerant (23) in the upper active regenerator (20) and compress the refrigerant (23) in the lower regenerator (20). The use of four three-way valves (36) eliminate the necessity of reversing the flow of the heat transfer fluid (35) through the system, insure counterflow fluid flows in both the cold heat exchanger and heat rejection exchanger at all times, and eliminate two heat exchangers as depicted in the embodiment of
The second half of the cycle is analogous to the first half of the cycle. In this instance the heat transfer fluid (35) flows through the upper regenerator (20) toward the bottom of the figure and is cooled by the refrigerant-filled tubes before passing through the cold heat exchanger (14) thermally coupled to the process stream heat exchanger (15) to accept heat from the process stream (11). The heat transfer fluid (35) then circulates through the lower regenerator (20), flowing toward the bottom of the figure, where it accepts heat from the refrigerant-filled tubes. This heat is expelled when the heat transfer fluid (35) passes through the heat rejection exchanger (13) that is thermally coupled to the heat sink exchanger (17). The valves (36) are then switched back to the configuration in
Various heat exchangers may be used for the AGRL. In one embodiment, small brazed-plate heat exchangers are used for the process stream heat exchangers. The heat rejection and heat sink exchangers may be compact plate-fin liquid-to-gas exchangers that have been used successfully in previous LNG liquefier designs and are available in the prior art. The duty of these exchangers is significantly more than the process stream heat exchangers. In the preferred embodiment, each AGRR stage rejects heat to a common water stream (the cooling fluid) that subsequently rejects heat to the environment through a water to air heat exchanger. The heat exchanger for removing excess heat from the cooling fluid can take the form of a finned tube with air cross flow as used in many air conditioners or similar vapor compression cycle refrigerators. These two operations are well defined and understood by those skilled in the art.
As mentioned, the preferred embodiment includes a cold box. Preferably, this cold box is vacuum insulated and of a simple design with a top plate for mounting the six stages of the AGRL. The assembly can be done easily with a crane to raise or lower the AGRL into the cold box. In this embodiment, the cold box is evacuated with a high vacuum turbo pump prior to operation. The cold box vacuum can be maintained by Cryopumping with zeolite containers attached to the cold end of one or the stages (e.g., an about 40 K stage because zeolite will adsorb helium at this temperature). Superinsulation is wrapped on all the cryogenic sections of the AGRL to reduce radiative heat leaks. All instrumentation can be mounted to the individual AGRR stages and the process stream may enter through vacuum feed throughs in the top plate of the cold box. The controls for the flows, the compressor/expanders, all valve drive motor controllers, and other operational components can be located outside the cold box.
In the preferred embodiment, the cryogenic liquid produced (i.e., LNG or LH2) is stored in a small vessel within the cold box with a double-walled vacuum jacketed or suitably insulated transfer line out of the vessel through the top plate for storage in an external cryogenic storage vessel. A level detector and or a mass flow meter may be placed in the vessel to directly measure the rate of liquefaction.
While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof have been shown by way of example in the drawings and herein described in detail. For example, although the refrigerant has been described throughout as helium gas, any fluid that can be compressed or expanded to cause a temperature change in the desired range may be used. Other mixed refrigerants or combined compressor/expander assemblies are also possible. One skilled in the art will recognize suitable substitutions. Similarly, the dimensions of the various components may be varied. The AGRL may be configured to liquefy other fluids. It should be understood, however, that it is not intended to limit the invention to the particular forms disclosed, but on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention as described.
The applicant claims priority from a Provisional Patent Application filed on Jul. 15, 2005, under Application No. 60/699,948.
The invention was created during a Phase I Small Business Innovation and Research award from NASA to CryoFuel Systems, Inc. under contract number NNJ04JC25C completed Jul. 15, 2004, under which the Government may have certain rights in this invention.