PULSE ELECTROTHERMAL DEICER FOR POWER CABLES RELATED APPLICATIONS
This application claims priority to U.S. Provisional Application Serial No. 60/545,038, filed 17 February 2004, and to U.S. Provisional Application Serial No. 60/497,442, filed 22 August 2003, each incorporated herein by reference. Commonly owned and copending U.S. Patent Application Serial No. 10/363,438 is also incorporated herein by reference.
BACKGROUND
Deicing is a process in which interfacial ice attached to a structure is either broken loose from the structure or melted. Some sort of external force (e.g., gravity, wind-drag, etc.) then removes the ice from the surface of the structure. Mechanical deicers require less energy than thermal deicers but they do not always remove all the ice from the structure being deiced. Also, the mechanical deicer may damage the structure being deiced, accelerating wear on the structure. Thermal deicing is effective but uses large amounts of energy (e.g., electricity) because of heat loss to convective heat exchange/absorption by bulk ice and by the structure being deiced. FIG. 1 A illustrates a cross section through a prior art aluminum conductor, steel reinforced (ACSR) cable 10. Cable 10 is typically used for power transmission. Cable 10 has seven steel wires 12 that add strength to cable 10, and thirty aluminum wires 14 that provide high current carrying capability (at a lower cost and weight as compared to steel) to cable 10. Cable 10 may also be constructed with fewer or more steel wires 12, and fewer or more aluminum wires 14, as a matter of design choice. FIG. IB illustrates a cross section through a prior art aluminum conductor, steel reinforced trapezoidal wire (ACSR/TW) cable 20. Cable 20 is also used for power transmission. Cable 20 has seven steel wires 22 that provide strength to cable 20, and thirty-six trapezoidal-shaped aluminum wires 24 that add high current carrying capability to cable 20 (at a lower cost and weight as compared to steel). Aluminum trapezoidal wires 24 are slightly more expensive to manufacture than aluminum wires 14, but reduce the overall diameter of cable 20 (as compared to cable
10) for a given cross-sectional aluminum area. Accordingly, cable 20 has less wind resistance and outer surface area as compared to cable 10; therefore, it also has less surface area for ice to adhere to than cable 10. Although use of cable 20 for power transmission results in less ice adhesion, the amount of ice that adheres to cable 20 is still sufficient to break cable 20 and, hence, disrupt power distribution. SUMMARY In one embodiment, a system deices a power cable that has an inner conductor and a metal shell. The metal shell is separated from the inner conductor by a dielectric material. A switch module is responsive to a control signal to divert power from the inner conductor to the metal shell, to generate heat to melt ice on the power cable. In another embodiment, a method deices a power cable having an inner conductor and a metal shell. A control signal is generated to indicate presence of ice on the power cable. Power is diverted from the inner conductor to the metal shell to generate heat to melt ice on the power cable. In another embodiment, a method prevents ice build-=up on a power cable. A power cable having an inner conductor and a metal shell is provided. Periodically, power is diverted from the inner conductor to the metal shell to generate heat to melt the ice. In another embodiment, a cable for power transmission includes an inner conductor, a dielectric layer, and an outer metal shell configured for resistive heating. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 A illustrates a cross section through an aluminum conductor, steel reinforced (ACSR) cable of the prior art. FIG. IB illustrates a cross section through an aluminum conductor, steel reinforced trapezoidal wire (ACSR/TW) cable of the prior art. FIG. 2 is a block diagram illustrating one system embodiment for deicing a surface. FIG. 3 is a graph illustrating temperature versus distance of heating power W for the exemplary embodiment of FIG. 2. FIG. 4 shows one pulse electrothermal deicer system embodiment.
FIG. 5 illustrates exemplary detail of a housing of one embodiment of an electrothermal deicer system connected with a power cable. FIG. 6 is a block diagram illustrating mechanical and electrical connections that support use of a switch in a pulse electrothermal deicer system. FIG. 7 shows exemplary deployment of pulse electrothermal deicing. FIG. 8A and FIG. 8B show cross sectional views of a power cable illustrating a shell with and without a longitudinal cut. FIG. 9A is a cross sectional view of an aluminum conductor, steel reinforced (ACSR) cable suitable for pulse electrothermal deicing. FIG. 9B is a cross sectional view of an aluminum conductor, steel reinforced trapezoidal wire (ACSR/TW) cable suitable for pulse electrothermal deicing. DETAILED DESCRIPTION OF THE DRAWINGS Systems and methods are described for deicing a surface, such as a surface of a power cable. The systems and methods described herein may advantageously employ, in certain embodiments, a low average power and a short active duration for deicing, thereby removing ice contamination in an effective manner. Certain of the following features may make such systems and methods superior to mechanical and conventional thermal deicing techniques. For example, in some cases, such features may reduce the energy taken to deice a cable by a factor of 100. Another advantage of one embodiment of the system is that it may use components, such as switches, that do not support a high voltage of a power transmission cable, but instead may be rated for much lower voltages, for example, a voltage rating that is a factor of 100 lower than the high voltage. A further advantage of such a system may be that it utilizes energy from the cable and does not, in this embodiment, employ a separate high power supply, high-power current transformer or high-power voltage transformer. FIG. 2 is a block diagram illustrating one electrothermal deicer system 100 for deicing a surface. A substrate 102 (e.g., an aircraft wing, refrigerator heat exchanger, car windshield, etc.) is shown with a layer of ice 104. A thin-film heater 106 is located at the interface of substrate 102 and ice 104. System 100 melts interfacial ice more economically than prior art deicing systems by reducing heat lost to both substrate 102 and ice 104. By placing thin-film heater 106 directly at the interface,
thermal resistance between thin-film heater 106 and ice 104 is reduced. Further, the heat lost from thin-film heater 106 to substrate 102 and to ice 104 is reduced by shortening the duration of the active heating time of thin-film heater 106. Thin-film heater 106 is shown connected to a controller 108 via a cable 110. Controller 108 operates to switch thin-film heater 106 on and off, for example. FIG. 3 is a graph 200 illustrating temperature T versus distance d of heating power W. Time t corresponds to heat diffusion time through ice 104 and substrate 102. Curve W1 represents a high density of heating power and curve W2 represents a low density of heating power. Temperature Tm represents the melting point of ice. The time t required for heat to diffuse through a thermally conductive substrate 102 over a distance L is given by the following equation: 2 (Eq. D t = - where D is a coefficient of heat diffusivity, defined in terms of a thermal conductivity coefficient λ, a substrate density p, and a substrate specific heat capacity c as: (Eq. 2) D = j-
When t is very short (e.g., a deicing pulse of short duration), the heat is temporarily encapsulated (captured) inside the substrate and the ice, rather than dissipating into the environment. Moreover, a shorter pulse of power heats a thinner layer of ice, thereby loosening the "grip" of the ice. Therefore, power is better conserved by focusing the power on the space where needed - at the interface. For the interface geometry shown in FIG. 2, the deicing pulse time tι required to warm up the interfacial ice from temperature T to a melting point T
m is:
where subscripts "i" refer to ice, subscripts "s" refer to the substrate materials, and W is a power per square meter. The total energy to heat ice to the melting point is then:
As is shown in Eq. 4, the higher the heating power W, the less energy it takes to start melting ice. A latent heat q for melting an interfacial ice layer of thickness d may be added to Eq. 4 to determine a power required to melt the ice in addition to the power required to reach the melting point. Accordingly, the total power is approximately: (τ
m ~ τ) (Eq. 5) Qt = iPi
ci
λi /vA -l + d
■ q - p
The minimum thickness d of the melted layer should be sufficient to enable the ice to slide on a viscous water film by action of gravity force, an air dragging force (such as in aviation), and/or a centrifugal force (such as from rotor-blades of helicopters and windmills). A typical value of a sufficient water film layer thickness in aviation is about 2 microns. Notice that the corresponding second term in Eq. 5 is usually less significant than the first term. Accordingly, the time predicted in Eq. 3 should not change much with the addition of the latent heat term. For a relatively thick heater film, a term originating from a thermal capacity of that film Q
h may be also added to Eq. 5:
where -4, C
h, and pi, are a heater material thickness, specific heat capacity and density, respectively. A maximum rms potential difference V across the dielectric is defined in terms of an rms current I and a shell resistance R: (Eq. 7) V = IR where R is defined in terms of a shell material resistivity ?, a section length L, and a shell thickness h: (Eq. 8) R = p-U(π-D-h) To save energy in deicing, a high instantaneous power may be used and applied in short energy pulses. The time between such pulses is defined by a rate of ice growth and tolerance to ice thickness. In one example, in aviation, ice thicker than about 3 mm may be removed. In transmission of electric energy, deicing pulses may
be applied either frequently to keep the substrate virtually free of ice, or only applied when an ice load becomes dangerous, such as on a power cable. FIG. 4 illustrates one pulse electrothermal deicer system 300, suitable to deice power cables. System 300 shows a power cable 301 which has an inner conductor 302, a metal shell 304 and a thin layer of dielectric material 306 separating metal shell 304 from inner conductor 302. System 300 also includes one or more switch modules 309, shown in FIG. 4 as 309(1) and 309(2). Each switch module 309 includes an electrical connection 311 between metal shell 304 and inner conductor 302, a switch 308 that interrupts current flow between adjacent sections of inner conductor 302, and a controller (not shown in FIG. 4 for clarity of illustration) for switching switch 308 on and off. System 300 may optionally include one or more ice detectors 312 (as described in more detail below). In an example of illustrative operation, to deice a section L of cable 301, an electric signal from the controller opens switch 308(1), as shown. When switch 308(1) opens, cable current flows from inner conductor 302 through electrical connection 311(1) to metal shell 304 of section L, and back to inner conductor 302 at electrical connection 311(2) and closed switch 308(2), thereby heating shell 304 along section L. Typically, one of switches 308 opens for a deicing pulse time of 0.1s to Is to deice a cable section (e.g., section L); but depending on electrical properties and weather conditions, the deicing pulse time can vary from about 1ms to 10 seconds. Dielectric material 306 separates inner conductor 302 from metal shell 304. Typically, dielectric material 306 is between 0.5mm to 2mm thick, to provide sufficient dielectric strength to withstand the potential difference developed between inner conductor 302 and metal shell 304. Dielectric material 306 can be a polymer, ceramic and/or other non-conductive material. A typical section (e.g., section L of cable 301) may exist as a part of a cable between two towers, for example, spanning 200m to 400m, or other lengths. Multiple sections may be deiced simultaneously, or sections may be deiced one by one in a "domino-like" manner (i.e., where sections are deiced sequentially). When sections are switched in and out of deicing mode for about Is, deicing may spread along cable 301 with a speed comparable with a speed of sound in air. Powering one section at a
time may reduce associated voltage and/or current fluctuations caused in the transmission cable by deicing. Sections may also be deiced in response to input from ice detectors (e.g., ice detector 312), which may lead to (a) some sections being deiced at the same time, and/or (b) some sections being deiced more frequently than other sections. It is not necessary to place a switch at each end of each section (e.g., section L of FIG. 4 and FIG. 6); rather, only one switch may be employed per section (e.g., switch 308(1) for section L in FIG. 4 and FIG. 6). Switches 308 of FIG. 4 may be electronic switches, electromechanical switches or combinations thereof. If a switch 308 fails, preferably it fails 'on' such that power transmission is not affected by a failure, and only deicing ability is lost; that is, it is undesirable to incur power loss caused by an inability to turn deicing off. Examples of electronic switches include thyristors, Insulated Gate Bipolar Transistors (IGBTs), high-current Metal Oxide Semiconductor Field Effect Transistors (MOSFETs), other power semiconductors, pre-packaged electronic switches and solid-state relays. Examples of electromechanical switches include relays and contactors, such as automobile starter coils. Combinations of both can be used to provide advantages of both. For example, a relatively small and inexpensive thyristor can interrupt a large current by switching at a zero crossing of the current; however, its voltage drop may cause large power dissipation in a steady state. A small and inexpensive relay connected in parallel provides low resistance and voltage drop during steady-state operation; but it may not interrupt the full current at full voltage. If the relay is turned off while the thyristor is on, the relay need not interrupt the full current at full voltage; the thyristor then conducts only for a very short time before it is turned off so its energy dissipation remains low. The thyristor may then turn off to produce a deicing pulse. After the deicing pulse, the thyristor turns on first, followed by the relay, to provide a similar advantage in the turn-on process. Mechanical switches (e.g., a relay or a contactor) may for example use gravity or a spring, such that contacts remain closed if the switch fails. FIG. 5 shows exemplary mechanical detail of a portion 600 of one switch module (e.g., switch module 309) of an electrothermal deicer system (e.g., system
300) connected with a power cable 301(3). Clamps 602(1) and 602(2) attach to ends of metal shell 304, and clamp 604 attaches to an end of inner conductor 302. Clamps 602 and 604 may be mechanically integrated into a single structure; a structure integrating clamps 602 and 604 may also include electrical connection therebetween (i.e., forming an electrical connection 311 as shown in FIG. 4). Clamps 602 and 604 attach to structural supports known as "dead ends" (not shown) that bear the tension present in cable 301(3), thus creating a region in which the tension is removed from current carrying conductors. The dead ends may be part of a housing that is configured inline with cable 301(3); in this configuration, outer conductor 606 may be configured as a cylinder and clamps 602 may be inner and outer concentric rings clamping outer conductor 606. Alternatively, dead ends may be configured through support from other structures (e.g., transmission towers) and outer conductor 606 may be configured to fit the geometry of the region between the dead ends. Thus, switches 308 (not shown in FIG. 5) and/or other components of switch modules 309 may be configured such that they need not bear the tension present in cable 301(3). FIG. 6 is a block diagram illustrating mechanical and electrical connections of a switch module 309(3) suitable for use in a pulse electrothermal deicing system 300. A housing 620 and an outer conductor 616 substantially surround a transformer 610, a controller 612, a switch 308(5) and connections therebetween. Outer conductor 616 and housing 620 may be mechanically and/or electrically separate, as shown, or may be integrated. Mechanical supports 614(1) and 614(2) clamp the layers of power cables 301(4) and 301(5) (for example, by using clamps 602 and 604 as shown in FIG. 8). Supports 614(1) and 614(2) are dead ends which bear the tension present in cables 301(4) and 301(5) respectively. In an inline configuration, mechanical supports 614(1) and 614(2) may be supported by housing 620 so that switch module 309(3) may be placed within a power line cable without external mechanical support (for example, as shown in FIG. 7). Alternatively, mechanical supports 614(1) and 614(2) may be supported by other support members such as transmission tower components (not shown). Inner conductors 618(1) and 618(2) (relieved of the tension present in cables 301(4) and 301(5) by mechanical supports 614) connect with switch 308(5). Current transformer 610 supplies power to controller 612 and/or switch
308(5). Since transformer 610 supplies only control power, its voltage and current output ratings may be much lower than the heating power delivered by deicer 300. FIG. 7 shows exemplary deployment of pulse electrothermal deicing to deice power cables 301(1) and 301(2). Power cables 301(1) and 301(2) are shown suspended from transmission towers 320(1), 320(2), 320(3) and 320(4). Switch module 309(4) controls pulse deicing of power cable 301(1) for section L; switch module 309(5) controls pulse deicing of power cable 301(2) for section L. Switch modules 309(6) and 309(7) control pulse deicing of adjacent sections L', L" of power cables 301(1) and 301(2), respectively. Switch modules 309(4) and 309(6) are located on, and therefore reference to a voltage of, power cable 301(1). Switch modules 309(5) and 309(7) are located on, and therefore reference to a voltage of, power cable 302(2). Accordingly, switches 308 within switch modules 309(4), 309(5), 309(6) and 309(7) do not need to be rated to the cable-to-cable voltage or cable-to-ground voltage of power cables 301(1) or 301(2); switches 308 may be, for example, rated for the voltage drop across a shell 304 of section L of power cable 301(1). In FIG. 7, two spans are included in section L; however, any number of towers and or spans may be included in section L. A single heating pulse may heat shell 304 of FIG. 4 to an ice melting point T
m of FIG. 3, thereby breaking ice bonds to cable 301. Gravitational force and cable vibrations may then remove ice from cable 301 before the melted interface refreezes. To enhance cable shaking, a longitudinal cut may be made in the outer shell as shown in FIG. 8A. FIGS. 8 A and 8B show cross sectional views of metal shell 304 of cable 301. FIG. 8A illustrates a cross section with no power applied and longitudinal cut 402 open; FIG. 8B illustrates a cross-section during a current pulse through metal shell 304, with longitudinal cut 402 closed (thereby "shaking" cable 301 and removing ice adhered thereto). Deicing system 300 may operate in two different modes, which may be implemented by controllers 612 of switch modules 309. Mode 1 - A preventative mode of operation includes short deicing pulses occurring periodically, often enough that percolation of drops does not occur.
Accordingly, a shell of ice does not lock or bond to the cable. Mode 1 is for example useful to prevent build-up on a cable so that, for example, ice is melted before it completely surrounds a power cable. Mode 2 - An emergency full-melting mode of operation provides higher power to melt ice by brute force. This allows removal of ice even if the ice has completely surrounded a cable such that it is linked and cannot fall off when adhesion is removed by a deicing pulse. Operation of system 300 in Mode 2 may be limited to an average power of, for example, 150 W so that cables 301 do not overheat (overheating may damage or melt the insulation, or could make the cable sag excessively). The average power may be achieved by adjusting the duration and frequency of deicing pulses, for example by pulse width modulation (PWM) or pulse frequency modulation (PFM). A section length (e.g, section L of FIG. 4 and FIG. 6) may be a distance between two towers (i.e., one span) or may be significantly shorter or longer. Examples are now described. Example #1 : One-span 400-m deicer. Deicer specifications: Cable diameter D is 35 mm Polyethylene dielectric layer thickness is 1mm Cable current I is 1000A Outside temperature T
m is -10°C Wind velocity v is 10 m/s Density of heating power Wis 60 kwatt/m2 (6.97 kwatt/m) The outer shell is made of pure aluminum of electrical resistivity ? = 2.5-10
"8ohm-m.
The shell thickness h is equal to: (Eq. 1-1) h = p-I2/(π-D)2W = 30 μm which is a typical thickness of adhesive aluminum tape, a readily and commercially available product. Density of pulse energy can be found from Eqs. 1-2 and 1-3 as : (Eq. 1-2) Q = 6.55 kVm2
A corresponding time duration is: (Eq. 1-3) -. --- 0.11 s
Notice, that an energy required to deice one meter of the cable is: (Eq. 1-4) Q' = Q-lm-π-D = 761 J If pulse deicing is done every hour, which is often quite sufficient, a corresponding "mean" power consumption per unit length of cable W is: (Eq. 1-5) W' = Q ' 3600s = 0.21 watt/m By comparison, a conventional continuous heater takes about 62 watt/m to keep the cable surface at 0°C under the same weather conditions (i.e., v = 10 m/s and T = -10°C). Thus, pulse deicer system 300 takes about 300 times less power to operate than a conventional heater. A voltage to de-ice a one-span section of a cable may be calculated as: (Eq. 1-6) V = I-Rsheii = W-L-π-D/I = 2.8 kV
This voltage may be reduced by using a thicker shell 304. Accordingly, for a 0.1mm thick shell, the voltage per span is .85 kV at 20 kwatt/m2 density of heating power. Example #2. A pulse electrothermal deicer for power cables was built and successfully tested in a cold room. An aluminum power cable of 35mm diameter was used. An insulating layer was made of two layers of duct-tape and an outer shell was made of 37 μm thick adhesive stainless steel foil. The deicer removed a dense snow pack frozen on its surface. FIG. 9A is a cross sectional view of an aluminum conductor, steel reinforced (ACSR) cable 500 suitable for use in pulse electrothermal deicing. Cable 500 has seven steel wires 502 that add strength to cable 500, and thirty aluminum wires 504 that add high current carrying capability to cable 500. Cable 500 also has a layer of smaller diameter aluminum wires 506 that are insulated from aluminum wires 504 by an insulating layer 508, thereby forming a metal shell around cable 500. Cable 500 may also be constructed with fewer or more steel wires 502, fewer or more aluminum wires 504, and fewer or more smaller diameter aluminum wires 506. Insulating layer 508 may, for example, be approximately 1mm thick and made from a polymer
material. The number and diameter of aluminum wires 506 are selected to provide the necessary heating power for cable 500. FIG. 9B is a cross sectional view of an aluminum conductor, steel reinforced trapezoidal wire (ACSR TW) cable 520 suitable for pulse electrothermal deicing. Cable 520 has seven steel wires 522 that add strength to cable 520, and twenty aluminum wires 524 that have trapezoidal cross-sections and add high current carrying capability to cable 520. Cable 520 also has a layer of smaller diameter aluminum wires 526, also with trapezoidal cross-sections, which are insulated from aluminum wires 524 by an insulating layer 528, thereby forming a metal shell around cable 520. Cable 520 may also be constructed with fewer or more steel wires 522, fewer or more aluminum wires 524, and fewer or more aluminum wires 526. Insulating layer 528 may, for example, be approximately 1mm thick, and made from a polymer. The number and diameter of aluminum wires 526 are selected to provide the necessary heating power for cable 520. As appreciated, there is little additional cost in manufacturing cables 500 and
520 compared to cables 10 and 20 of FIG. 1A and IB, respectively, thus adding little cost to implementing system 300. Very large amounts of power are available for melting ice using system 300, making system 300 suitable for severe weather conditions. Power for heating, for switches 308 and for control electronics 612 is taken from the power cable, so no additional power sources are needed. Switches 308, transformers 610 and control electronics 612 may have voltage ratings based on the voltage drop across metal shell 304 for the length of section L, are not referenced to ground voltage; they therefore are not exposed to a full power cable voltage, which is typically between 100 kN and 1000 kN. Thus, switches 308, current transformers 610 and control electronics 612 may be significantly smaller and cheaper than components that would be necessary to switch the full power cable voltage. Also, the inductance between metal shell 304 and inner conductor 302 is small, such that inductive spikes on switch 308 are reduced. Only one switch 308 is typically used per section L; additional cables (known as bundles) that conduct the same phase across section L may also be switched simultaneously by switch 308, to reduce the number of switches employed.