CROSS REFERENCE TO RELATED APPLICATIONS
This application is related to co-pending U.S. patent application Ser. No. 12/758,674 (attorney docket number 080077) filed by William H. Scofield and entitled “Electronic System Cooler”, commonly assigned with this application and incorporated herein by reference in its entirety.
- Top of Page
This application is directed, in general, to cooling of electronic systems.
An electronic system dissipates heat that, unless removed, increases the temperature of components within the system. If allowed to rise too high, the temperature may reduce the operating life of some components, and in some cases may result in loss of a service provided by the electronic system. A cooling system may thus be used to reduce the temperature of the electronic system.
- Top of Page
One embodiment provides a cooling system. The cooling system includes an electronic enclosure and a liquid-cooled heat exchanger configured to receive heated air from within the enclosure. A geothermal loop is connected to the heat exchanger to form a closed-loop path. The closed-loop path is suitable for circulating a pressurized liquid refrigerant between the geothermal loop and the heat exchanger to transfer heat from the heated air to the geothermal loop without the liquid refrigerant undergoing a phase change.
Another embodiment provides a method. The method includes configuring a liquid-cooled heat exchanger to receive heated air from an electronic enclosure. The heat exchanger is connected to a geothermal loop to form a closed-loop path. The closed-loop path is suitable for circulating a pressurized liquid refrigerant between the geothermal loop and the heat exchanger to transfer heat from the heated air to the geothermal loop without the liquid refrigerant undergoing a phase change.
Reference is now made to the following descriptions taken in conjunction with the accompanying drawings, in which:
FIG. 1 illustrates an embodiment of a system including an electronic enclosure and a cooling system having a closed-loop path;
FIG. 2 illustrates a vapor pressure curve representative of a refrigerant that may be used as a liquid refrigerant within the cooling system of FIG. 1; and
FIGS. 3A and 3B present a method that may be used to fabricate a cooling system, e.g., the cooling system of FIG. 1.
- Top of Page
Some electronic system installations are located at a site that is difficult to reach, is off the electric grid, or both. An example of such a system is a remote cellular tower transceiver system. In some conventional installations, the electronic system is cooled by a refrigeration system, including a compressor, evaporator and condenser, and a refrigerant that is compressed and expanded in a refrigeration cycle. The power required to operate such a conventional system may be substantial. For example, a refrigeration system that dissipates 10 kW may be needed to remove 9 KW of heat dissipated by the electronic system. Thus, the site may require electrical infrastructure sufficient to provide 19-20 kW to the combined electronic and refrigeration systems.
The electronic system may be situated in a relatively inaccessible location, so the cost of repairing the electronic system may be quite expensive, and the loss of service supported by the electronic system, such as a cell phone tower, may result in lost revenues and customer goodwill. Moreover, the conventional refrigeration system typically includes various components that must be serviced, and may fail unpredictably, adding potential sources of failure and expense. Repairing and maintaining the refrigeration system may involve expensive or difficult travel and access by a service provider.
In other conventional installations, a geothermal cooling system using water or a water/glycol solution has been used. However, in such systems the coolant poses a significant risk to electronics being cooled. If a leak develops near these components, aqueous coolant may spill onto electronic assemblies or power supplies. An aqueous coolant may have a low enough resistivity, e.g. from dissolved solids (salts), to conduct a significant current. The current may cause immediate failure of electronic components due to shorting or arcing. Even if no immediate failure results, ionic contaminants may react with various parts of the electronic, such as tin-based solder connections, causing corrosion that may eventually lead to failure.
The inventor has recognized that the deficiencies of conventional practice may be reduced or overcome by a cooling system based on geothermal cooling using a liquid refrigerant that is circulated substantially without a phase change. The liquid refrigerant is expected to substantially reduce or eliminate the aforementioned risks of shorting and corrosion. Moreover, the geothermal cooling system is expected to have an inherently greater reliability than a cooling system that relies on refrigeration, e.g. a compression-expansion cycle. Furthermore, the power required to operate the geothermal cooling system is expected to be significantly lower than that needed for a refrigeration system. Thus, expenses associated with the equipment site may be significantly reduced, and reliability of the electronic system increased.
Turning to FIG. 1, illustrated is an embodiment of a system 100. The system 100 includes an electronic system 105 and a cooling system 110 configured to circulate a liquid refrigerant as a coolant. Herein the terms “liquid refrigerant” and “coolant” may be used interchangeably. Aspects of the refrigerant are described further below. The electronic system 105 includes an enclosure 115 that includes electrical components 120 that dissipate heat when operating. A fan 123 may circulate air within the enclosure through a liquid-cooled heat exchanger 125 having an inlet 126 and an outlet 127. Herein, a “liquid-cooled heat exchanger” is a heat exchanger configured to transfer heat from a warmer source to a cooler liquid coolant. The heat exchanger 125 is configured to receive heated air circulating within the enclosure 115, and to transfer heat from the heated air to a liquid coolant flowing in a closed-loop path that includes the heat exchanger 125. In various embodiments the heat exchanger 125 is a micro-channel heat exchanger.
The cooling system 110 includes the heat exchanger 125 and various components configured to transport heat therefrom. A first coolant line 130 connects a pump 135 to the heat exchanger 125. The pump has an inlet 136 and an outlet 137. A second coolant line 140 connects the heat exchanger 125 to a geothermal loop 145 having an inlet 146 and an outlet 147. A third coolant line 150 connects the geothermal loop 145 to the pump 135. The heat exchanger 125, pump 135, geothermal loop 145, first line 130, second line 140, and third line 150 form a closed-loop path configured to circulate the liquid refrigerant.
Herein and in the claims, the term “closed-loop path” refers to an assembly of components configured to circulate a liquid refrigerant between the heat exchanger 125 and the geothermal loop 145 without a compression-expansion cycle and having no breaks in the path that would allow the liquid refrigerant to collect or escape under normal operating conditions. In various embodiments the refrigerant is under pressure to maintain a liquid phase.
The geothermal loop 145 is distinguished from a conventional geothermal loop configured for an aqueous coolant by being configured to maintain the pressure of the liquid refrigerant therein. Thus, the geothermal loop may include components sometimes used in refrigeration systems, including high-pressure tubing and fittings and/or pressure seals.
The geothermal loop 145 is illustrated as including a first vertical loop 145a and a second vertical loop 145b. The geothermal loop 145 is not limited to any particular number of loops, however. In other embodiments, the geothermal loop 145 may be configured to transfer heat to a body of water. In such cases, the geothermal loop 145 may be any other appropriate orientation with respect to the water surface, including about parallel thereto.
In various embodiments, the geothermal loop 145 is located below a surface 160 of the earth. “Below a surface of the earth” encompasses embodiments in which the geothermal loop 145 is embedded in soil, e.g., buried in a shaft, and embodiments in which the geothermal loop 145 is underwater, e.g., submerged in a body of water such as a river, lake or pond. For brevity of discussion, for embodiments in which the geothermal loop 145 is located below a surface of the earth, the geothermal loop 145 are referred to herein and in the claims as “buried.”
When the geothermal loop 145 is long enough, a portion of the geothermal loop 145 may be located in an isothermal region of the ground. Such a region may be characterized as having a temperature that is relatively insensitive to seasonal fluctuation. In some cases, it is believed that the isothermal region will have a relatively constant temperature of about 10-15° C. Thus the ground at a sufficient depth may provide an effective heat sink for the coolant.
The coolant circulates through the heat exchanger 125 where heat is transferred from warm air therein to the coolant, thereby cooling the air. The warmed coolant flows to the geothermal loop 145, in which the heat carried by the coolant is transferred to the relatively cool ground (or water). The coolant returns to the pump 135 via the inlet 136. The head pressure is sufficient to circulate the coolant. However, unlike a refrigeration compressor, the pump 135 is not configured to compress the refrigerant from the gas phase to the liquid phase.
The coolant flows to the geothermal loop 145 via the second line 140. While it is generally expected that the coolant will remain in the liquid phase within the second line 140, the presence of some vapor is within the scope of the disclosure. Such vapor, if present, is expected to be transient and occupy a negligible volume of the closed-loop path. Any such vapor is expected to condense before or within the geothermal loop 145.
The pressure differential between the outlet 137 and the inlet 136 of the pump 135 should be sufficient to overcome drag between the coolant and the various paths through which the coolant passes, and to lift the coolant through any vertical paths in the cooling system 110. For example, a height H between the bottom of the geothermal loop 145 and the inlet 136 may be 100 meters or greater.
An optional receiver 155 located inline between the geothermal loop 145 and the pump 135 may be used to trap any vapor mixed with the coolant to prevent loss of pumping action by the pump 135. In some embodiments a bubble detection system (not shown) may be used in lieu of or in addition to the receiver 155. Such a system may disable the pump 135 when bubbles are detected near the inlet 136 of the pump 135. The bubble detection system may then re-enable the pump 135 when the closed-loop path is determined to be free of bubbles.
The coolant may be any conventional or nonconventional liquid refrigerant. Such substances are typically not significantly electrically conducting. Thus such a coolant is expected to significantly reduce the risk to electrical components 120 posed by a coolant leak. As described further below, an electrically nonconductive coolant may also have other properties beneficial in various cases.
As used herein, electrically nonconductive means that the resistivity of the coolant is at least about 1 MΩ-cm. In various applications it may be preferable that the resistivity is at least about 5 MΩ-cm to ensure fewer ionic contaminants. In some cases it may be more preferable that the resistivity is at least about 10 MΩ-cm. For reference, ultrapure deionized (DI) water may have a maximum resistivity of about 18 MΩ-cm.
The coolant may include any chemical compound suitable for use as a refrigerant, such as chlorocarbons, fluorocarbons, hydro-chloro-fluorocarbons (HCFCs), and hydro-fluorocarbons (HFCs). The coolant may be a pure refrigerant or may include a mixture of two or more refrigerants. The coolant may also include other chemical components such as lubricating oil or a corrosion inhibiter that may be advantageous in the operation of the cooling system 110.
In various embodiments the coolant is pressurized. Herein and in the claims, a coolant is pressurized when the coolant has a boiling point less than about 50° C. and is maintained at a pressure greater than the vapor pressure of the coolant at about 50° C. The vapor pressure is typically greater at higher temperature and lesser at lower temperature, so the pressure within the closed-loop may vary with location. In some cases the coolant vapor pressure may be greater than about 600 kPa at standard temperature (about 25° C.). At operating temperature (e.g. about 50° C.) such coolants may be pressurized at 1 MPa or greater.
The refrigerant may be placed in the closed-loop path via a fill valve 141. The fill valve 141 may include, e.g. a Schrader-type valve mechanism. The fill valve 141 may be located in any accessible portion of the closed-loop path. In some cases it may be desirable to evacuate the closed-loop path before charging the closed-loop path with the refrigerant.
In some embodiments, the coolant is a non-ozone-depleting refrigerant, such as those that do not include chlorine. Non-ozone depleting refrigerants may be characterized as having an ozone depletion potential (ODP) of about zero. An example chlorine-free refrigerant is 1,1,1,2,-tetrafluoroethane (CH2F—CF3) , often referred to as R134a, having a boiling temperature Tb of about −26° C. at standard pressure.
In some other embodiments, the refrigerant has a low global warming potential (GWP). Those skilled in the pertinent art will appreciate that a refrigerant may be characterized by a GWP that represents the potential for that chemical to contribute to global warming, relative to the effect of CO2. Thus, for example, the GWP of CO2 is unity. One nonlimiting example of a refrigerant that is chorine-free and has a low GWP is 2,3,3,3-tetrafluoroprop-1-ene (CH2═CFCF3), sometimes referred to as HFO-1234yf. HFO-1234yf or equivalent is available from various manufacturers, including Honeywell International, Inc, Morristown, N.J. 07962, USA. HFO-1234yf has a boiling temperature Tb of about −29° C. at standard pressure, and is regarded as having a GWP of about 10 or less, e.g. between about 4 and about 6.
In some cases it may be desirable to select a coolant with a low viscosity. For the purpose of this discussion a coolant has a low viscosity when it has a viscosity of about 250 μPa·s or less at about 25° C. For reference, water has a viscosity of about 900 μPa·s at about 25° C. It is expected that a coolant with a lower viscosity will require a less powerful pump 135 to circulate the coolant than would a coolant with a higher viscosity due to flow-related friction. Thus, a low-viscosity coolant may require less energy to circulate than does water. Moreover, a low-viscosity coolant may be particularly advantageous when used with a micro-channel heat exchanger to provide a greater flow rate through coolant passages therein. In a nonlimiting example, R134a is a low-viscosity coolant, having a reported viscosity of about 15 μPa·s at 25° C. HFO-1234yf and similar refrigerants are also expected to be low-viscosity coolants.
The use of a coolant with a low boiling temperature provides a significant advantage with respect to conventional liquid coolants. As previously described, if a leak develops within the heat exchanger 125, the coolant may leak onto electronics components within the enclosure 115. In the case of an aqueous coolant, even if the coolant is substantially free of dissolved ions, the coolant may cause the electronic components to corrode. However, when the coolant has a low boiling point, e.g. no greater than about 10° C., then the coolant will rapidly evaporate. When the boiling point is no greater than about −20° C., the coolant may evaporate before reaching the electronic components, substantially eliminating the risk to the electronic components. Even if the coolant leak causes the system 100 to shut down, the cost to repair the system 100 should be substantially less than if electronic components were damaged as well.
The pressure of the coolant at the outlet 147 is expected to be reduced in part by the weight of the coolant column within the geothermal loop 145a. So that the coolant does not vaporize near the outlet 147, the coolant may be advantageously maintained at a sufficient pressure so that the minimum pressure at the outlet 147 remains above the vapor pressure of the coolant at the temperature at that location.
The highest temperature of the system 110 is expected to be near the outlet 127 of the heat exchanger 125. Thus the vapor pressure of the coolant is expected to be greatest at this location. To prevent vaporization of the coolant the minimum pressure at the outlet 127 should be greater than the vapor pressure of the coolant at the outlet temperature.
To illustrate these points, FIG. 2 presents a vapor pressure versus temperature characteristic representative of both R134a and HFO-1234yf. The vapor pressure is about 250 kPa at 15° C., about 670 kPa at 25° C., and about 1.3 MPa at 50° C.
If, for example, R134a is used as the coolant, and the temperature thereof is about 50° C. at the outlet 127, a minimum pressure of about 1.3 MPa is expected to prevent the coolant from vaporizing. On the other hand, at the outlet 147, the temperature may be only about 15° C. At this point only about 250 kPa is needed to prevent vaporization of the coolant. Some embodiments may thus include a pressure within the closed-loop path that is sufficient to ensure that the pressure at the outlet 147 during operation of the cooling system 110 remains above 250 kPa and the pressure at the outlet 127 remains above 1.3 MPa.
In some embodiments a pressure reducer (not shown) may be used to separate a high pressure region from a low-pressure region of the closed-loop path to facilitate maintaining a desired pressure distribution within the closed-loop system. The pressure reducer may be controllable by a controller 170 to adjust for various pressures and/or heat loads from the electronic system 105. The controller 170 is described in greater detail below.
Circulating a refrigerant without undergoing a phase change is markedly different than conventional use of refrigerants. Typically, a refrigerant is used in a compression-expansion cycle that transports heat in part by the phase changes that occur upon compression (liquification) and expansion (vaporization). In the case of various embodiments of the disclosure, vaporization of the coolant is either prevented or negligible. In either case, heat transfer from the heat exchanger 125 to the geothermal loop 145 is substantially by virtue of the heat capacity of the liquid coolant, rather than condensation of a vapor phase of the coolant.
Those skilled in the pertinent art will appreciate that the thermodynamics of the cooling system 110 will depend in part on the thermal properties of the coolant. In particular, the heat capacity and thermal conductivity of the coolant are expected to have a significant role in this regard. The heat capacity and thermal conductivity of a liquid refrigerant are expected to be lower than that of an aqueous coolant. Thus in various embodiments employing a liquid refrigerant the flow rate of the refrigerant may be greater relative to an aqueous coolant to reflect the lower heat transport rate provided by a unit of the liquid refrigerant. However, such an increase of flow rate may be tempered by the lower thermal conductivity associated with some liquid refrigerants relative to aqueous coolants. In practice an acceptable flow rate may be empirically determined by one skilled in the pertinent art for a particular system without undue experimentation.
In various embodiments, the pump 135 produces a pressure sufficient to overcome a pressure between the coolant at the inlet 126 and the outlet 127 of the heat exchanger 125. For example, in the previous example of R134A, the vapor pressure of R134A is about 0.25 MPa at 15° C. and about 1.3 MPa at about 50° C. Thus, the pump 135 may need to produce a pressure of at least about 1 MPa to move the coolant from the inlet 126 to the outlet 127. For a coolant with a higher boiling point, the pressure differential produced by the pump 135 may be more modest, e.g. adequate to overcome friction of the coolant circulating in the closed-loop path as previously described.
Relative to a conventional refrigerated remote equipment installation, embodiments of the disclosure are expected to consume far less power. For example, as previously described a conventional system may require 10 kW to remove 9 kW of dissipated power. In contrast, embodiments of the disclosure are expected to consume only about 500-1000 W to cool a 9 kW load. The reduced total power load of the site in turn reduces the cost of delivery of electrical power to the site, and maintenance costs of the delivery system. Thus, embodiments of the disclosure provide a significant advantage over conventional cooling approaches.
Transport of heat from the coolant to the environment, e.g., air above-ground, or soil or water below ground, may be enhanced by the use of an optional flow turbulence generator 165. The flow turbulence generator 165 may take the form of one or more of a vortex generator, a zig-zag or spiral portion of the second line 140, or one or more protrusions attached to the inner wall of the second line 140 configured to disrupt smooth, e.g. laminar, flow of the coolant without imposing significant back pressure on the flow of the coolant. Turbulence produced by the flow turbulence generator 165 is expected to increase mixing of the coolant, thereby increasing efficiency of heat transfer to soil, water or air in contact with the second line 140. The turbulence generator 165 may be most effective with a low-viscosity coolant such as R134A or HFO-1234yf.
Optionally, a control system including the flow controller 170 is configured to control the pump 135 and/or an optional flow valve 173 to control the flow rate of coolant through the heat exchanger 125. Flow control may serve at least two purposes.
First, the flow may be reduced when the heat load produced by the electronic system 105 falls, e.g. because of reduced loading on a function provided by the electronic system 105. Reduced coolant flow may be advantageous in some cases to maintain consistent pressure characteristics within the closed-loop path.
Second, flow control may be used to ensure that the temperature of the coolant within the heat exchanger 125 does not fall below a dew point of air within the enclosure 115. The controller 170 may receive signals from sensors within the heat exchanger 125 that determine temperature and relative humidity. The controller 170 may determine therefrom the dew point of the air, and control the flow of coolant to maintain a temperature of the coolant greater than the dew point. Such operation may be advantageous to reduce or eliminate condensation formed within the heat exchanger 125 that might otherwise cause corrosion or electrical shorting of components within the electronic system 105.
In some embodiments the cooling system includes an auxiliary air-cooled heat exchanger 175. Herein, an “air-cooled heat exchanger” is a heat exchanger configured to transfer heat from a warmer liquid coolant to cooler air in contact with the air-cooled heat exchanger. The heat exchanger 175 is configured to transfer heat from the coolant to cooler air near the system 100. Coolant flows to the heat exchanger 175 via the second line 140 in parallel with, or alternative to, the geothermal loop 145. When the heat exchanger 175 is present in the cooling system 110 the coolant path including the heat exchanger 175 is considered part of the closed-loop coolant path. Coolant may flow to the heat exchanger 175 via a directional valve 180, and may reenter the line 150 via a directional valve 185. The valves 180, 185 may be controlled, e.g. by the controller 170 which may sense ambient air temperature near the cooling system 110. When the ambient temperature falls below a value at which air cooling becomes more efficient or effective than, e.g., ground cooling, the controller may control the valves 180, 185 to route coolant through the heat exchanger 175. The heat exchanger 175 may optionally include a fan to move air thereover. Thus, for example, the cooling system 110 may rely on air cooling in winter months and ground cooling in summer months.
Other cooling methods may be combined with the geothermal cooling of the cooling system 110. In some embodiments, heat-radiating fins may be attached to the second line 140 to aid the transfer of the heat carried by the coolant to surrounding the surrounding air. In other embodiments a water jacket may be used to cool the coolant within the second line 140.
Turning to FIG. 3A, a method 300 is shown that may be employed, e.g. to manufacture the system 100. Without limitation the method 300 is described using elements of the system 100 for illustration. The steps of the method 300 may be performed in the order presented or in another order.
In a step 310, a liquid-cooled heat exchanger, e.g. the heat exchanger 125, is configured to receive air from an electronic enclosure, e.g., the enclosure 115.
In a step 320 the heat exchanger is connected to a geothermal loop to form a closed-loop path. The closed-loop path is suitable for circulating a pressurized liquid refrigerant between the geothermal loop and the heat exchanger to transfer heat from the air to the geothermal loop without the liquid refrigerant undergoing a phase change.
FIG. 3B presents optional steps of the method 300.
In a step 330 the closed-loop path is filled with a liquid refrigerant as previously described. Filling the closed-loop path may include placing the refrigerant under sufficient pressure such that the closed-loop path is substantially filled with the liquid phase of the refrigerant. Substantially filled means the liquid phase of the refrigerant occupies at least about 90%, preferably at least about 95%, and more preferably at least about 99% of the volume of the closed-loop path at standard temperature.
In a step 340 the control system 170 is configured to control the rate of coolant flow through the heat exchanger to maintain a temperature of the coolant within the heat exchanger above a dew point of the air within the enclosure 115.
In a step 350 the air-cooled heat exchanger 175 is configured to receive the coolant and to transfer heat from the coolant to air outside enclosure 115.
Optionally the coolant is a liquid fluorocarbon, such as R134a or HFO-1234yf. Optionally, the coolant has a boiling point at standard pressure less than about −20° C. Optionally the coolant is a refrigerant having a global warming potential of about 10 or less. Optionally, the coolant has a viscosity less than about 250 μPa·s.
Those skilled in the art to which this application relates will appreciate that other and further additions, deletions, substitutions and modifications may be made to the described embodiments.