1、4744 (RP-1074) Determination of Dielectric Properties of Ref rig erants Andrew Michael Gbur Associate Member ASHRAE ABSTRACT The electrical properties of dielectric constant, dissi- pation factor, and breakdown voltage were determined for refrigerants 11, 12, 22, 32, 123, 124, 125, 134a, 143a, 236fa
2、, 245fa, 404A, 407C, 410A, 507A, 508A, and508B. Allmeasure- ments were mude at 25C. The measurements of dielectric constant und dissipation factor were obtained for saturated liquid und vapor as well as 1 atmosphere vapol: Breakdown voltages were determined for only the vapor phase at various pressu
3、res and electrode spacings (gap). The breakdown in the saturated liquid occurs by slight electric heating of the satu- rated liquid, causing it to vaporize, which then is simple break- down in saturated vapor. In addition, a breakdown voltage is lower for vapor than for compressed liquid, and the lo
4、west breakdown voltage is of concern, ASTMD924,0150, andD2477 apparatus and testproce- dures were utilized in this work. In addition, data from the published literature are presented for comparison. The justi- $cation for this work is to produce comparative data for the refrigerants in this study, u
5、tilizing the same testing apparatus and conditions. INTRODUCTION Refrigerants come in direct contact with electrical motor components and electrically live conductors in hermetically sealed refrigeration and air-conditioning systems. Here the electric motor can be cooled by liquid refrigerant and is
6、 submerged in refrigerant vapor. The electrical properties of the refrigerants are important to motor performance and reliabil- ity. Electrical properties of interest include breakdown volt- age, dielectric constant, resistance, and power factor (ac dissipation factor). The dielectric strength of a
7、medium is the breakdown voltage (BDV) divided by the distance between the electrodes at breakdown (arcing). Paschens law states that the breakdown voltage is a function of the product of gap width times the pressure. Paschen curves are the graphical represen- tation of this function. The dielectric
8、strength is not to be confused with the equally important property of dielectric constant. The dielectric constant is the ability of a material to shield a charge from surrounding charges. The dielectric constant of a material will determine the capacitance of a system for a given geometry. The powe
9、r factor is a measure of the electrical losses in the system. Both the dielectric constant and the power factor describe the performance change of a motor, from operating in air to operating in refrigerant. A refrigerant can differ significantly from air for these proper- ties. Dielectric property d
10、ata for refrigerants are scarce and incomplete. This work presents measurements to compare with the values that currently exist for the mature refrigerants and provide additional electrical property data for the new refrigerants on the market. As the trend progresses toward the use of HCFC and HFC r
11、efrigerants and mixtures thereof, the need for a comprehensive resource on the electrical properties of these refrigerants becomes necessary. Currently, systems appear to be in proper and safe working order, but what is not known is what type of future modifications can be tolerated in the system an
12、d how these changes would affect the safety factors and reliability. If a manufacturer wants to provide smaller and more efficient compressors and motors, they will need to decrease the spacing between live electrical parts and the dimensions of the outer shells. Will such spacing be sufi- cient to
13、provide electrical protection? If the refrigerant in the system is changed, has the likelihood of electrical shockkhort Andy Gbur is general manager at Interteks Refrigeration Laboratory, Columbus, Ohio. 26 02005 ASHRAE. increased? This project is designed to assist in answering these questions and
14、provide data that are not currently available to the designers of these units (air conditioning, refiigeration, and heat pumping). The electrical characterization of refrigerants will help to avoid repeating history, i.e., the unexplainable Hi Pot failures upon R-22 introduction. The Hi Pot test is
15、a high-voltage potential motor test. The motor is subjected to twice the volt- age rating plus 1000 volts for one minute. Dielectric break- down during this time is considered a failure. When R-22 was first utilized, the system designers failed to account for the lower insulating properies compared
16、to those of the CFC refrigerants. Thereby, when a compressor was constructed using similar gaps and spacing, the possibility of shorts and electric discharges was increased. The data presented will help ensure reliable units and prediction against electrical losses. The data are needed to check moto
17、r design for efficiency and possible breakdown failure in usage and the High Pot motor test. DIELECTRIC CONSTANT AND DISSIPATION FACTOR MEASUREMENTS The following measurements were taken for refrigerants 11, 12,22,32, 123, 124, 125, 134a, 143a, 236fa, 245fa, 404A, 407C, 410A, 507A, 508A, and 508B: a
18、. Vapor dielectric constant and dissipation factor at 25C and pressures of 101.325 kPa and the saturation pressure. b. Liquid dielectric constant and dissipation factor at 25C and saturation pressure. Test Apparatus The electrical measurements of dielectric constant and dissipation factor at 1 O00 H
19、z were made using a Model 2500A capacitance and dissipation factor test set manufactured by Andeen-Hagerline. These measurements conformed to ASTM D150, Test Methods for A-C Loss Characteristics and Permittivity (Dielectric Constant) of Solid Electrical Insulat- ing Materials. The equipment was cali
20、brated such that the results bear a known and valid relationship, and the equipment was validated prior to use against RC dissipation factor stan- dards. The test cell was constructed as shown in Figure Al .3 of ASTM D924-92, Standard Test Method for Dissipation Factor (or Power Factor) and Relative
21、 Permittivity (Dielectric Constant) of Electrical Insulating Liquids, except that modi- fications were necessary to accommodate pressures from vacuum to saturation pressure at 25C for the refrigerants of interest. This cell was a three-terminal cell manufactured by Tettex Instruments and was their P
22、ilot Cell for liquid insula- tion materials Model No. 2905, 15 mL volume (Figure 1). The test cell temperature was not regulated, but the appa- ratus was located in a humidity-controlled laboratory whose temperature was maintained at 25k2“C with forced air venti- lation. The cell was allowed suffici
23、ent time to equilibrate with the ambient, which was confirmed using NIST traceable ther- mometers as required by ASTM D924 and ASTM D2477. This ensured that cell handling, filling, and dielectric heating processes did not cause deviations from the required temper- ature. All temperature-indicating d
24、evices were calibrated in accordance with ANSI Z540/ISO Guide 25, Technical Competence of Calibration and Testing Laboratories. Experimental Procedure The following procedure was employed for all tests measuring for the dielectric constant and dissipation factor. Figure 1 Tettex capacitance cell use
25、d for dielectric 1,5 Insulation 2, 8-12 Connections, stainless steel Connection for high voltage Guard electrode, stainless steel High-voltage electrode, stainless steel Low voltage electrode, stainless steel constant und dissipation factor determination. 13 Socket ASHRAE Transactions: Research 27 1
26、. 2. 3. 4. 5. 6. 7. 8. 9. The Tettex cell (see Figure 1) was prepared by cleaning with spectrophotometric grade acetone and drying with a warm nitrogen stream and vacuum. The cell was placed into the high-pressure chamber and the electrical connections were attached. All cables were shielded and a g
27、uard electrode was employed. The chamber was sealed and evacuated to less than 150 microns Hg. When vacuum was achieved, the vacuum pump was valved off and the chamber was filled to 1 atmosphere from a cylin- der of dry air (lo ppm H20). The capacitance and dissipation factor were recorded. This was
28、 used as a verification to ensure that all connections were complete and the equipment maintained its calibra- tion. The chamber was again evacuated to less than 150 microns Hg. The refngerant of interest was attached to the chamber via a stainless steel line. The hose was purged to remove any air t
29、rapped in the line. The chamber was filled with refrigerant to 1 atm. The sample was then allowed to equilibrate for 30 minutes inside the test cell and then five measurements were made for each property. The chamber was then filled to the saturation pressure of the gas at 25OC. 10. The sample was a
30、gain allowed to equilibrate for 30 minutes inside the test cell and again five measurements were made for each property. 1 1. The refrigerant source cylinder was then inverted and liquid refngerant was forced into the chamber via slight heating of the source cylinder. 12. The sample was then allowed
31、 to equilibrate for 60 minutes before five measurements were made again for capacitance and dissipation factor. BREAKDOWN VOLTAGE MEASUREMENTS For breakdown voltage testing, the test cell was constructed as shown in Figure 1 according to ASTM D2477- 96, Standard Test Method for the Dielectric Breakd
32、own Volt- age and Dielectric Strength of Insulating Gases and Commer- cial Power Frequencies, except that modifications were made as necessary to accommodate pressures from vacuum to satu- ration pressure at 25OC for the refrigerants of interest (see Figure 2). A vernier mechanism in the breakdown v
33、oltage test cell provided the gap spacings from 0.000 to 8.000 mm (0.000 to 0.315 in.) with 95% confidence interval of 0.025 mm (0.001 in.) in the gap setting. The vernier mechanism was zeroed prior to setting the gap for each refrigerant. The electrodes were inspected at regular intervals to ensure
34、 that pitting did not invalidate the dimension or shape requirements specified in ASTM D 2477-96, which could lead to a distorted electrical field. The test cell was shielded from light and other radiation sources to decrease the dark current and increase the sensitiv- ity of detecting electrical br
35、eakdown of the refrigerants. The test cell temperature was regulated by conducting all measurements in a temperature and humidity controlled labo- ratory, which was maintained at 25C with forced air ventila- tion. The cell was allowed sufficient time to equilibrate to the ambient temperature and was
36、 fitted with NIST traceable ther- mometers as required by ASTM D924 and ASTM D2477 to ensure that cell handling, filling, and dielectric heating processes did not cause deviations from the required temper- ature. All temperatures were measured using indicating devices, which were calibrated in accor
37、dance with ANSI Z54OlISO Guide 25, Technical Competence of Calibration and Testing Laboratories. The test cell pressure was controlled and monitored at all times. For tests conducted at one atmosphere, a mercury column was employed to ensure that a standard atmosphere of 760 mm Hg was used instead o
38、f prevailing local conditions. For tests conducted at other pres- sures, an appropriately sized Bourdon tube test gauge with an Figure 2 Dielectric breakdown voltage cell speciJied in ASTM D 24 77. 28 ASH RAE Transactions: Research accuracy of 1% was used. These readings were adjusted, as necessary,
39、 to correct the results according to the current baro- metric pressure in the laboratory. All pressure-indicating devices were calibrated in accordance with ANSI Z54O/ISO Guide 25, Technical Competence of Calibration and Testing Laboratories. Test Apparatus breakdown voltage is as follows: Dielectri
40、c breakdown voltage tester: High-voltage Model DTS-60D, O to 60 kV output, Arc Detection with 50-millisecond shutdown. Fluke Model 12 Multimeter, Fluke Models 80k-40 and 80k-6 high-voltage probes. Dielectric cell manufactured to ASTM D 2477 specifi- cations with a 304 stainless steel body in order t
41、o obtain measurements on high-pressure fluids. A second dielec- tric cell made to ASTM D 2477 specifications with a glass body, allowing visual observations for measure- ments not exceeding 6.44 bar (80 psig). The specific equipment used in the measurement of the Breakdown voltage measurements were
42、made at 61.8 Hz, as the selected test equipment employs this frequency, and the results were adequately representative of the refrigerants performance at power line frequencies. Experimental Procedure measuring for the dielectric breakdown voltage (BDV). The following procedure was employed for all
43、tests 1. 2. 3. 4. 5. 6. The BDV cell was prepared by cleaning with spectropho- tometric grade acetone and drying with a warm nitrogen stream under vacuum. A new stainless steel ball was used for each new refrigerant tested. The brass disc was polished clear of any char and free of pitting (see Notes
44、 1 and 2 below). The cell was then evacuated to less than 150 microns Hg. A cylinder containing the refrigerant under test was attached to the cell and the lines were purged to remove traces of air. The cell was pressurized to the desired test pressure with refngerant (see Note 3) and the gap distan
45、ce between the ball and plate was adjusted to the desired value. The voltage was applied via the breakdown voltage tester, and upon breakdown the voltage automatically shut off and the breakdown voltage was recorded. A Fluke multimeter equipped with high-voltage probes confirmed the results of the i
46、nstrument. Additional refrigerant was added to achieve the next Paschen curve data point. The addition of refngerant provided a dilution for any ions that may have been formed during the discharge. The very low pressures were produced by filling the chamber with vapor, which was produced by an expan
47、sion of liquid refrigerant. If the pres- sure was greater than desired, it was adjusted by evacuating the excess. Steps 5 and 6 were repeated until a plot of the breakdown voltage vs. pressure x distance produced a minimum data point. Additional data points were determined above and below the minima
48、 for confirmation. The entire procedure was then repeated with the same refrigerant to validate the first set of data. Note 1: The required stainless steel ball attached to the end of the source electrode was replaced prior to initiating subsequent tests on other refrigerants. This avoided the addi-
49、 tive charring and pitting on the surface of the component and avoided any dimensional changes to the test cell. Note 2: The brass disc referred to above was a perma- nent receptor electrode and therefore could not be replaced. It is essential that this component be thoroughly and completely cleaned and polished between all tests. Note 3: All refrigerants used in testing (dielectric and breakdown measurements) were transferred to sealed cylin- ders where all noncondensable gases were removed by freez- ing the refrigerants in liquid nitrogen and pulling vacuum on the cylinder to approx
