Network manageable advanced gas sensor apparatus and method
09851277 · 2017-12-26
Assignee
Inventors
Cpc classification
G01M3/26
PHYSICS
International classification
G01M3/26
PHYSICS
Abstract
Mechanical, electronic, algorithmic, and computer network facets are combined to create a highly integrated advanced gas sensor. A sensor is integrated into switchgear housings. These sensors integrated into high voltage switchgear products, deployed by electric utility end users in replacement and expansion cycles, function to detect and mitigate atmospheric pollution caused by leaking SF.sub.6. As its associated gas insulated tank is charged with 10 to 350 lbs. of SF.sub.6, each gas sensor monitors its local cache of gas, accurately sensing and computing fractional percentage losses (emissions) and gains (maintenance replacement) in SF.sub.6 mass, storing data in onboard data logs, and communicating data when triggered by detection events or in response to remote requests over a hierarchical communications network, a process that continues without labor until a fractional leak is automatically detected and reported creating the opportunity for early leak mitigation.
Claims
1. A method for computing gas density comprising the steps of: acquiring a raw temperature measurement from a temperature probe in connection with a gas; converting said raw temperature measurement to an accurate temperature reading using pre-determined temperature calibration information; acquiring a raw displacement measurement from a displacement sensor in connection with a moveable element responding to the pressure and to the temperature of said gas; converting said raw displacement measurement to an accurate displacement measurement using pre-determined displacement calibration information; determining a nominal pressure measurement corresponding to said accurate displacement measurement using pre-determined displacement to pressure transformation information; converting said nominal pressure measurement to an accurate pressure reading using said accurate temperature reading in conjunction with pre-determined temperature to pressure compensation information, and computing the density of said gas using said accurate temperature reading and said accurate pressure reading in conjunction with a gas law.
2. A method for computing gas density as claimed in claim 1 wherein said temperature probe is a thermistor.
3. A method for computing gas density as claimed in claim 1 wherein said displacement sensor is a reflective object sensor.
4. A method for computing gas density as claimed in claim 1 wherein said displacement sensor is a Hall effect sensor.
5. A method for computing gas density as claimed in claim 1 wherein said gas is sulfur hexafluoride.
6. A method for computing gas density as claimed in claim 1 where said gas law is a second or higher order polynomial function of said gas density and said polynomial function coefficients are a function of said accurate temperature reading.
7. A method for computing gas density as claimed in claim 1 further comprising the step of: recording all measurements, readings, and calculations in a microcontroller memory including said raw temperature measurement, said accurate temperature reading, said raw displacement measurement, said nominal pressure measurement, said accurate pressure reading, and said gas density.
8. A method for computing gas density as claimed in claim 7 further comprising the steps of: repeating said foregoing steps periodically thereby creating a stored time sequence of said measurements, readings, and gas density, and utilizing said stored time sequence to determine a leakage of gas or an addition of gas.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
(21)
(22)
(23)
(24)
(25)
DESCRIPTION OF THE INVENTION
(26) As stated above, the instant invention supports a 100-fold reduction in gas emissions. What does this imply for gas density measurement requirements? To address this question, begin by considering that, at a temperature of 20° C., the operating pressure for the breakers of table 1 ranges from 64 psig to 82 psig, a span of 18 psi.
(27) Table 1 also gives the nominal gas mass change attributable to pressure change for each breaker under the aforementioned isothermal conditions. The function is simply proportional to the differential tank volume of the various breakers given the isothermal assumption. As expected, the largest tank represents a worst case requirement for mass sensing resolution since smaller pressure changes accompany larger gas losses (large mass changes). In general, larger tanks will require higher resolution measurements to detect unit changes in gas mass.
(28) TABLE-US-00001 TABLE 1 Breaker Model Designation HS HP-1 HP-2 HP-3 HPI-1 HPI-2 Rated 72.5 145 169 242 345 550 Maximum Voltage (kV) Interrupting 31.5 40 40 40 362 40 Current Rating (kA) Tank Volume 0.151 0.561 0.732 1.171 3.367 3.542 (cubic meters) SF6 weight at 5.2 19.1 24.9 39.9 114.8 120.7 fill pressure (kg) SF6 weight at 4.7 17.4 22.7 36.3 104.3 109.8 nominal (kg) SF6 weight at 4.3 16.0 20.9 33.4 96.1 101.1 alarm (kg) SF6 weight at 4.0 14.8 19.3 31.0 89.0 93.6 lockout (kg) SF6 Emission 0.8 3.1 4.1 6.5 18.7 19.7 between fill and alarm (kg) SF6 mass per 0.063 0.232 0.302 0.484 1.391 1.464 unit pressure (kg/psi) Distribution 50% 14% 13% 15% 6% 2% frequency [12]
(29) Table 1—Representative gas insulated breakers with OEM recommended SF6 fill conditions. If filled to just below fill capacity, breaker type HPI-2 would emit 19.7 kg of SF6 before the alarm threshold would trigger. Typically the so-called “nameplate capacity” will be three times larger than the above tank capacity since the breaker comprises three phases each with its individual tank The approximate frequency with which various sizes occur in practice is attributable from Blackman.
(30) Now the question arises, what is the magnitude of gas loss one needs to begin detecting? SF6 emission rates studied by various methods to date appear to place gas emissions in the range of 5% to 10% of total nameplate capacity annually. Accounting for frequency of distribution of breakers by voltage rating (and therefore by tank size), the weighted average of the nameplate capacities is approximately 73 kg (remember—3 tanks per breaker typically). In a study of 2,329 breakers by Blackman, 170 (7.3%) were found to be leaking. The amount of gas emitted to atmosphere annually may thereby be estimated at 3.7 kg to 7.3 kg per breaker (5% to 10% of 73kg). The actual leaks arise from the aforementioned 7.3% of the breaker population. Therefore, the average leakage amount per leaking breaker is on the order of 50 kg to 100 kg annually.
(31) The sensor-gas interface mechanism as one component of the advanced gas sensor has many important aspects. The use a mechanical bellows approach is utilized for several reasons. These reasons include the bellow's simplicity, reliability, and broad use in SF.sub.6 gas density switch applications. The use of a mechanical bellows leads to a requirement for detecting and processing mechanical displacement information. Processing the displacement information supports accurate gas pressure inferences.
(32) The advanced gas sensor combines a bellows sensing element with an MCU Electronics module comprising electronics and software for acquiring raw displacement and temperature information and processing these into accurate measurements.
(33) Reliable pressure and temperature readings must ultimately be rendered from raw sensor data. The present invention utilizes an efficient signal processing chain for this purpose. Noise, stability, and other potential problems are thereby identified and overcome.
(34) Processed pressure and temperature readings must be interpreted to predict gas density which in turn predicts gas mass changes in light of known, rigid tank volumes. The process, in isothermal conditions, is relatively straight forward. Under conditions of changing temperature, the process becomes more challenging. Ideal gas law and virial equations with alternative techniques for calculating temperature dependent coefficients form the foundations of the algorithms utilized for this purpose.
(35) As stated earlier, the present invention uses a mechanical bellows approach to gas interface and pressure sensing. The advantages of this choice are described above. Mechanical bellows components are readily available from a variety of sources including Solon Manufacturing of Chardon, OH. Mechanical bellows are widely used in mechanical, gas density monitoring products that enjoy a dominant share of the North American alarm-monitoring market.
(36) The bellows expands under increasing pressure. In the configuration of the embodiments set forth herein, the bellows actuates a rigid coupling to a platen. The platen's starting position and translational gain are simultaneously adjusted with a counter-biasing coil spring. Nominal gain in the range of 0.001″ platen deflection per 1 psi change is typically achieved.
(37) In the mechanical density monitor application, the platen carries bi-metal elements that in turn actuate snap-action micro-switches under conditions of sufficient displacement. The bi-metal elements provide a mechanical temperature compensation mechanism.
(38) Contrastingly, in the instant invention, the platen is adapted to carry displacement sensor components which take the form of reflective surfaces, magnets, and other components supporting displacement detection alternatives.
(39) As stated above, the invention targets supporting a 100-fold reduction in gas emissions. In the discussion above, it was deduced that a 100-fold improvement in emissions mitigation implies gas sensor sensitivities of 0.5 kg to 1.0 kg. According to table 1 above, this suggests a differential pressure resolution on the order of 16 psi to 0.35 psi. Recall that the operating span of interest is approximately 18 psi. Thus the required pressure measurement resolution (before correction) is in the range of 1 part in 1.2 to 1 part in 52.7. In digital measurement terms, this corresponds to a 1 bit to 6 bit dynamic range, which is achieved using a microcontroller and 12 bit analog to digital converters.
(40) Core bellows devices, prior to any modification, have been bench tested for displacement response over the pressure range of interest at 20° C. Conventional gauge room equipment was used to measure displacement. Regulated compressed air provided pressure actuation. Pressure gradients in both directions have been utilized to quantify hysteresis, and several runs are made to assess short-term repeatability. Analysis of data captured in these tests was analyzed and definitively demonstrates the bellows fitness for the application in this invention as shown hereinbelow.
(41) A microcontroller 1301 is used to perform displacement sensing and temperature sensing, and to communicate raw data to the other controller functions (via asynchronous serial communications initially).
(42) MCU subsystem modularity allows easy substitution of alternative circuits for the powered by battery DISPLACEMENT INTERFACE 1309 and HI-RES TRANSDUCER 1310. The MCU 1301 monitors battery state of charge and computes circuit power consumption as well, an important distinguishing characteristic of circuit and algorithmic alternatives. The FLASH memory based MCU 1301 may be conveniently reprogrammed to adapt to varying sensing requirements. Operating parameters may be programmed and acquired data retrieved over the bidirectional, asynchronous communications interface.
(43) Initial choices for HI-RES TRANSDUCER 1310 used to measure platen 601 displacement include Hall Effect and photo diode/transistor technologies. The optical alternatives comprise both transmission and reflective technologies. Piezo strain gauge and ultrasonic systems are possible as well.
(44) Processing and calibration requirements for rendering accurate pressure readings from displacement data are included. Temperature channels are also logged during operation. All data generated by the displacement and temperature measurement blocks is forwarded to the Temperature and Pressure Processing blocks. The invention covers the operating and temperature ranges of interest. Temperature gradient and leak rate tests are also satisfied. The invention includes the signal processing necessary for rendering reliable pressure and temperature readings from raw displacement and temperature sensor data.
(45)
(46)
(47)
(48) N.B. Calibration is achieved entirely using digital techniques to determine coefficients stored onboard in nonvolatile memory. Use of precision or adjustable components is avoided in favor of standard tolerance, inexpensive, high stability components.
(49) An equation of state model is required to compute the target gas density from calibrated temperature and pressure data. A first order Gas Density Model is used and provides satisfactory results in many cases.
(50) The well known virial form set forth below as equation 2 utilizing coefficient functions for SF.sub.6 selected from various perspectives is an alternative embodiment:
(51)
(52) Where p, V, n, R, and T have their usual meanings in the ideal gas law, and B(T) and C(T) are the second and third virial coefficients respectively, each non-linear functions of temperature T.
(53) This step further draws upon recent work by Scalabrin describing a computationally efficient neural network technique for computing coefficients in a certain form of state equation.
(54) It is an important aspect of the instant invention to use a micro-power microcontroller platform to sense gas density to sufficient accuracy to discern 0.5 kg emission events under a range of conditions of interest for the largest tank volumes expected.
(55)
(56)
(57)
(58)
(59) Referring to
(60) One important object of the invention is to determine if gas is being loss from the switchgear. The gas sensor apparatus operates over a wide range of temperature and pressure conditions other than standard temperature and pressure conditions. If pressure of the gas rises, but the mass of the gas within a known volume stays the same (ie no loss occurs), then the apparent pressure in the volume (tank) appears to increase. The bimetallic strip 708, however, adds a downward force on lever 601 to counteract the additional force of the gas within the bellows due to an increase in gas temperature. If pressure of the gas decreases, but the mass of the gas within a known volumes stays the same (ie no loss of gas occurs), then the apparent pressure in the volume (tank) appears to decrease. In a similar manner, an apparent decrease in gas pressure due to a relatively low temperature, is compensated by an upward force on lever 601 to counteract the reduction in force of the gas within the bellows due to a decrease in gas temperature.
(61) A magnet is affixed to the lever 601. A reflective surface is also affixed to the lever 601. A hall effect sensor is applied to the switch printed circuit board 501. A reflective object sensor is affixed to the switch printed circuit board. In
(62)
(63) Referring to
(64)
(65) Where P is gas pressure in the system, V is the volume of gas which is fixed by the equipment's rigid tank, R is a constant, T is temperature, and n is the mass quantity of gas. With V and R constant, measuring P and T determines n/V, the gas density.
(66) The gas sensor apparatus includes switch actuator elements 704 which reside on lever 601 which engage the actuators 618AA of snap action switches 618A-D as illustrated in
(67)
(68)
(69)
(70) Referring to
(71)
(72)
(73)
(74)
(75)
(76) The ideal gas law restated:
(77)
(78) p=absolute pressure (pounds per square inch or psi)
(79) V=volume (cubic meters)
(80) T=temperature (Kelvin)
(81) n=gas quantity in moles (mol)
(82) R=gas constant=1.2095×10.sup.−3
(83) It should be noted that, p, the pressure in (1) is the absolute pressure (reference to a vacuum) which differs by atmospheric or barometric pressure from the pressure indicated by a typical gauge in atmospheric conditions. This can be stated mathematically as:
p=p.sub.abs=p.sub.g+p.sub.atm (3)
(84) p=p.sub.abs=absolute pressure (psi)
(85) pg=gauge pressure (psi)
(86) p.sub.atm=atmospheric pressure (psi)
(87) With n the gas quantity in mol known, the mass quantity for a particular gas is derived from its molar weight:
m=nM.sub.m (4)
(88) m=gas quantity in grams (g)
(89) n=gas quantity in moles (mol)
(90) M.sub.m=molar mass of gas species (g/mol)
(91) A sequence of measurements of gas mass m.sub.i=m.sub.1, m.sub.2, . . . m.sub.j can be derived using corresponding sequences of pressure p.sub.i and temperature T.sub.i measurements given only that the volume V, atmospheric pressure p.sub.atm, and gauge pressure p.sub.g corresponding to each point in the sequence are known. A change in gas mass foretells a leak when a measurement m.sub.j is less than a measurement m.sub.k made sometime earlier (k<j). Conversely, the addition of gas is detected when m.sub.j is greater than m.sub.k. In a non-leaking system, all of the m.sub.i will be substantially equal.
(92) Acquiring temperature sequence T.sub.i begins by microcontroller 1301 using analog to digital converter 1303 applied to temperature interface 1307 accessing temperature probes 1308A through 1308D to acquire raw sensor measurements. Raw sensor measurements are then converted to accurate temperature readings through a calibration process such as that depicted in
(93) Acquiring pressure sequence pi is somewhat more involved. It begins again with microcontroller 1301 using analog to digital converter 1303 applied to displacement interface 1309 accessing high resolution displacement transducer 1310 to acquire raw displacement sensor measurements. Unlike temperature measurements, there is no simple transformation of raw displacement measurements to absolute pressure, however. Firstly, a raw displacement sensor measurement is utilized by the MCU to compute calibrated lever displacement dimension according to calibration data such as that depicted in
(94) Once calibrated lever displacement dimension is derived, initial gauge pressure estimate can be computed using secondary calibration data as depicted in
(95) The intrinsic temperature compensation of the lever system comprising bi-metal hinge 708 must now be taken into account. In the absence of the bi-metal element, lever position would simply track temperature variations. For the fixed volume V, gas pressure increases proportional to increasing temperature (and vice versa). With only the resistance of coil spring 608, lever dimension 1004 would increase proportionately with the varying force exerted by bellows 1003. The bi-metal element is conceived to neutralize this temperature induced pressure variation. As temperature increases, the bi-metal exerts approximately equal magnitude equal force directed oppositely to the increased upward force of the bellows with the approximate result that the lever dimension remains constant. The converse occurs as temperature decreases. These mechanics alone allow the mechanism to operate as a low resolution density monitor wherein eventual changes in lever position represent approximate changes in gas mass (as opposed to pressure variations due to temperature), and, for fixed volume V, gas density. With the advent of the microcontroller in the present invention, it is possible to improve accuracy and flexibility of gas monitoring including the electronic measurement of pressure, temperature, gas content, and gas density as explained above.
(96) To complete the derivation of absolute gas pressure p from displacement and temperature sensor measurements, the initial gas gauge pressure estimate as above must itself be compensated for the temperature behavior introduced by the bi-metal element. The appropriate compensation is derived from the data in
(97) To complete the example, given a typical tank volume V of 1 cubic meter, along with a molar mass for SF.sub.6 gas of 146.055 g/mol, the gas mass m.sub.i is computed according to (1) and (4) to be 38.62 kg. The entire process is implemented by microcontroller 1301 in combination with the electronic elements of
REFERENCE NUMERALS
(98) 100 perspective view of sensor 101 cover 102 liquid tight pushbutton 103 cover retaining screw 104 housing 105 sensor connector 105A sensor connector contact pin 105B connector nut 106 manifold block 106A manifold block bolt hole 106B first gas port 107 display deadfront 107A deadfront gasket edge 200 perspective view of the sensor with cover and housing removed 202 pushbutton cable 203 pushbutton cable connector 204 processor PCB pushbutton connector 205 processor PCB 204 processor PCB pushbutton connector 205 processor PCB 300 perspective view of the sensor with pushbutton removed 301 display PCB 302 display PCB retaining screw 303 display PCB standoff 304 display digit 305 display PCB processor PCB connector 306 display PCB coil spring clearance hole 204 processor PCB pushbutton connector 205 processor PCB 400 perspective view of the sensor with display PCB removed 401 processor PCB retaining screw 402 processor PCB standoff 403 processor PCB coil spring clearance hole 404 processor PCB display PCB connector 405 processor PCB switch PCB connector 105 sensor connector 105A sensor connector contact pin 500 perspective view of the sensor with processor PCB removed 501 switch PCB 502 switch PCB connector PCB 503 switch PCB flexible circuit element 504 switch PCB retaining screw 505 switch PCB processor PCB connector 506 switch connections 507 thermistor connections 507A thermistor 507B thermistor stalk 508 reflective object sensor connection 509 switch PCB coil spring and switch adjustment clearance cutout 507A first thermistor stalk 507B first thermistor 507C third thermistor stalk 507D third thermistor 600 left side view of the sensor internal components 600A detail of lever and displacement mechanisms 600B right side view of the sensor internal components 600A detail of lever and displacement mechanisms 601 lever 602 base plate 603 riser 604 adapter flange 605 second gas port 606 coil spring nut 607 coil spring washer 608 coil spring 601 lever 609 Hall effect sensor 610 reflective object sensor 611 magnet 612 magnet boss 613 reflective surface 614 reflective surface boss 615 ball 616 ball spring adjuster 617A third thermistor stalk 617B third thermistor 617C fourth thermistor stalk 617D fourth thermistor 618A first switch 618B second switch 618C third switch 618D fourth switch 619 bimetal hinge base retaining plate 620 bimetal hinge base retaining plate nut 700 top view of the sensor lever and switch actuator elements 701 coil spring stud 702 switch actuator element adjuster screws 703 unused actuator element adjuster screw threaded hole 704 switch actuator elements 705 switch actuator elements flange screw 706 bimetal hinge lever retaining plate 707 bimetal hinge lever retaining plate nut 708 bimetal hinge 709A first switch PCB mounting boss 709B second switch PCB mounting boss 709C third switch PCB mounting boss 709D fourth switch PCB mounting boss 800 bottom view of switch PCB 800A perspective view of switch PCB from bottom 900A perspective view of sensor internal components from bottom 1000 front view of sensor internal components 1001 switch actuator element adjuster screw boss 1000A front view of sensor internal components with riser cutaway 1000B front view of lever displacement detail 1002 bellows lever coupling 1003 bellows 1004 lever displacement dimension 1004 lever displacement dimension 1100 top view of sensor internal components 1100A crosssection view from right side of sensor internal components 1101 gas port 1102 sensor module 1103 processor module 1104 stabilizer 1200 top view of sensor internal components 1200A crosssection view from front of sensor internal components 1300 hardware block diagram 1301 MCU (microcontroller unit) 1302 communication subsystem 1303 analog to digital converter subsystem 1304 digital to analog converter subsystem 1305 digital I/O interface subsystem 1306 safety limit detection subsystem 1307 temperature interface 1308A first temperature sensor 1308B second temperature sensor 1308C third temperature sensor 1308D fourth temperature sensor 1309 displacement transducer interface subsystem 1310 high resolution displacement transducer 1311 battery 1312 network management controller 1400 processing block diagram 1401 raw conversions from displacement and temperature subsystems 1402 signal to noise enhancement 1403 displacement to pressure calculation 1404 pressure calibration 1405 digital filtering subsystem 1406 signal processing parameter set 1407 pressure readings output to gas mass calculations and user interfaces 1500 lever position as a function of gas pressure at 25 C 1501 lever response from 0 psig to 60 psig, ball nose spring plunger operating 1502 lever response from 60 to 100 psig, coil spring and bimetal hinge only 1503 a particular lever position of 0.026″ 1504 a pressure of 81 psig corresponds to position of 0.026″ 1600 normalized sensor response as a function of lever position 1601 response of reflective object sensor 1602 response of Hall effect sensor 1603 a particular HES sensor response of 0.55 1604 a particular lever position of 0.026″ corresponds to sensor response of 0.55 1700 graph of pressure compensation required as a function of temperature 1701 pressure compensation required as a function of temperature 1702 zero compensation required at reference temperature 1703 a particular temperature 1704 a particular pressure compensation corresponds to a particular temperature 1800 graph of temperature as a function of normalized temperature sensor response