Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2021 Aug 12;6(33):21534–21542. doi: 10.1021/acsomega.1c02512

Exploring the Decomposition Products of 1,3,3,3-Tetrafluoropropene and Perfluoro-(3-methylbutan-2-one) Gas Mixtures in Medium-Voltage Electrical Switchgear as Alternatives to SF6

María Luz Alonso †,*, Rosa María Alonso †, José Ignacio Lombraña ‡, Jesús Izcara §, Josu Izagirre §
PMCID: PMC8388090  PMID: 34471756

Abstract

graphic file with name ao1c02512_0010.jpg

In this work, binary and ternary gas mixtures of 1,3,3,3-tetrafluoropropene, HFO-1234ze(E), and perfluoro-(3-methylbutan-2-one), CF3C(O)CF(CF3)2, with CO2 and synthetic air, are presented as alternatives to SF6 in medium-voltage electrical equipment. They were used in four medium-voltage switchgear cubicles replacing SF6 gas, and after a period of time, under permanent 30 kV AC voltage, gas mixture samples were extracted and analyzed on the same day using a validated methodology based on gas chromatography (GC) coupled to mass spectrometry (MS) and thermal conductivity (TCD). CF4 (tetrafluoromethane), C2F6 (hexafluoroethane), C3F6 (hexafluoropropylene), C3HF7 (1,1,1,2,2,3,3-heptafluoropropane), CH2F2 (difluoromethane), and the cis and trans-C3H2F4 (1,3,3,3-tetrafluoropropene) have been identified as decomposition products in these gas mixtures. In addition, a quantity of water has been observed, as well as CO in one of the cubicles. The most abundant decomposition products identified in gas mixture samples (C3HF7 and C3F6) together with water and CO content have been quantified using commercial gas mixture reference standards. The toxicity and global warming of the analyzed compounds are evaluated to determine the most adequate gas mixture among those studied as a candidate to substitute SF6.

1. Introduction

Nowadays, manufacturers of medium-voltage switchgear frequently use sulfur hexafluoride (SF6) in the equipment because of its very good insulation and switching properties. SF6 has unique characteristics such as relatively low toxicity, extreme inertness, and high dielectric strength. However, SF6 is one of the most potent greenhouse gases that cause significant global warming, 23.500 times higher than CO2, according to IPCC’s Fifth Assessment Report (AR5).1−3

Although there are other insulating technologies (oil or solid insulation), they present drawbacks because of safety or cost issues, making gas insulation the best choice for distribution switchgears. Therefore, there is a need for a gaseous alternative, that is, gas or gas mixture, which can replace SF6 in medium-voltage electrical equipment. Alternative gas or gas mixtures must have an acceptable combination of low boiling point, chemical stability, nonflammability, high dielectric strength, and as low as possible global warming potential (GWP), at least less than 150, based on the Annex III of Regulation (EU) No. 517/2014 of the European Parliament and of the Council of 16 April 2014. This regulation has been established for fluorinated greenhouse gases used in foams, refrigerators, and freezers that contain HFCs (hermetically sealed equipment), movable room air-conditioning equipment or technical aerosols.

In addition, alternative gas mixtures must have relatively low toxicity, and depending on the specific application, they should be functional at temperatures as low as −25 °C or even as low as −30 °C.3,4

Investigations have been focused on natural gases (N2, CO2, and dry air) and more recently on new halogenated products with low GWP such as trifluoroiodomethane (CF3I), perfluorinated ketones, hydrofluoroolefins, and fluoronitriles.5−8

Ormazabal Corporate Technology has found that gas mixtures with hydrofluoroolefins such as 1,3,3,3-tetrafluoropropene HFO-1234ze(E) (HFO) and perfluoroketones such as perfluoro-(3-methylbutan-2-one) CF3C(O)CF(CF3)2 (C5K) can be environmentally and economically feasible for their use in medium-voltage electrical distribution switchgear (from 12 to 40.5 kV), reaching similar values of electrical stiffness to that of SF6. Unfortunately, these polyfluorinated gases have a boiling point higher than SF6, Table 1, which is a drawback for their use in low-temperature applications of medium-voltage switchgear. However, the addition of dry air, nitrogen, or carbon dioxide in the gas mixtures reduces the boiling points of their mixtures, and they can also inhibit the formation of some decomposition products.

Table 1. Physicochemical Properties of HFO and C5K in Comparison with SF6.

1.

Open in a new tab

HFO-1234zeE has good dielectric strength properties, as can be seen in Table 1, and its GWP is less than 1, according to IPCC Fifth Assessment Report [1]. This value is better than that of SF6 (GWP of 23.500). The insulation properties of HFO-1234ze(E) are close to those of SF6 and can probably be used in its pure state for replacing SF6 in medium-voltage applications.4,5

Perfluoroketones can be used as an additive to natural gases such as N2, CO2, or dry air, but their concentration in the mixture will be limited by the minimal operation temperature of the electrical switchgear-specific application.

Some perfluoroketones such as C5K are substantially nontoxic in the pure state and present high insulation capabilities, in particular, a high dielectric strength (or breakdown field strength), and, at the same time, an extremely low GWP (see Table 1). Indeed, by adding fluorine in the chemical formulation of perfluoroketones, the dielectric strength of the complex molecule is improved. In contrast, the boiling point increases considerably.5−9

The toxicity of these gases (collected in Table 1) is defined as the acute toxicity (LC50) and chronic toxicity. Low values of LC50 correspond to high gas toxicity. HFO and C5K have high LC50 values, corresponding to category 4 (very low acute toxicity) for HFO (as well as SF6) and category 5 (gas to be taken care of) for C5K, according to the National Institute of Workers.6 Chronic toxicity is characterized by the threshold limit values time-weighted average (TLV-TWA). The TLV-TWA must be higher than the maximum value that can occur permanently in a factory in normal conditions because of normal leakage during gas handling. It should be higher than 50 ppmv in order to be used in switchgear factories.9 HFO and C5K satisfy this factor.

The study of the behavior of HFO and C5K mixtures and their possible decomposition products at 30 kV in medium-voltage electrical switchgear requires the development of a suitable analytical method for their monitorization. Gas chromatography (GC) coupled to mass spectrometry (MS), flame ionization (FID), or thermal conductivity (TCD) detectors3,5−15 have been the analytical techniques frequently chosen for this purpose. Two-dimensional (2D) chromatography has been preferentially used to perform an optimal separation of the chromatographic peaks, but good resolution has not been obtained.

Therefore, the aim of this work is to study the chemical stability of different gas mixtures of dielectric gases, such as HFO and C5K, in combination with CO2 or synthetic air, as alternative gases to SF6, at 30 kV in medium-voltage electrical switchgear. In order to carry out this study, a simple and rapid one-dimensional (1D) chromatographic method with an optimal chromatographic resolution has been developed and validated for the quantitative analysis of gas mixtures with HFO and C5K and their possible decomposition products.

2. Results and Discussion

2.1. Chromatographic Behavior of HFO and C5K Gases

In Figure 1, the GC/MS and GC/TCD chromatograms of HFO and C5K reference gases, obtained under the conditions optimized are shown. Table 2 shows the retention times (tr) and characteristic ions (m/z) of the studied compounds.

Figure 1.

Figure 1

Open in a new tab

GC/MS and GC/TCD chromatograms of 40% v/v HFO in synthetic air and 20% v/v C5K in synthetic air.

Table 2. Retention Times (tr) and Characteristic m/z Ions of the Studied Compounds Obtained by GC/MS and GC/TCD Chromatographic Systemsa.

2.1.

Open in a new tab
a

Ions in bold are used for quantification.

2.2. Validation of the Chromatographic System

Once chromatographic conditions were optimized to achieve a good chromatographic resolution of the two reference gases for both chromatographic systems used (GC/MS and GC/TCD), validation of the analytical method was carried out in terms of selectivity, limit of detection (LOD) and quantification (LOQ), linear concentration range, and intermediate precision.

2.2.1. Selectivity

The selectivity parameter determines whether a chromatographic peak is only due to the analyte of interest and not to the presence of interferences in the sample. As can be seen in Figure 2, chromatographic peaks do not appear in the blank sample (synthetic air) at the retention times of the two analytes (Table 3).

Figure 2.

Figure 2

Open in a new tab

Synthetic air chromatograms obtained by GC/MS (a) and by GC/TCD systems (b).

Table 3. LOD and LOQ Values (% v/v) and Variation of Response with C5K and HFO Concentrations for GC/MS and GC/TCD Systems.

2.2.1.

Open in a new tab

2.2.2. Limit of Detection and Quantification

The lowest concentration of the analyte that can be detected using the method at a specified level of confidence is named the LOD. The LOD refers to the qualitative aspect of the method but does not refer to whether the analyte can be reliably quantified. For this purpose, the LOQ is additionally required. The LOQ is defined as the minimum amount of analyte that can be quantified with an acceptable level of precision and accuracy. A signal/noise ratio (S/N) was used for the LOD and LOQ. A S/N value of 3 for LOD and S/N of 10 for LOQ are acceptable.

2.2.3. Linear Concentration Range

The ability of an analytical method to give responses that are proportional to the amounts of an analyte in a sample determines its linear concentration range. Different gas mixtures composed of different percentages of C5K and HFO with synthetic air were prepared in a stainless-steel cylinder in order to create different calibration standards. These calibration standards were stored in Tedlar bags until analysis. In Table 3, the linear regression equations obtained when the chromatographic peak area and gas concentration data were treated for GC/MS and GC/TCD are collected. The linear concentration range obtained was from LOQ to 20% v/v for C5K and LOQ to 40% v/v for HFO.

2.2.4. Intermediate Precision

Two types of precision were determined: intraday analyses in working sessions of 6 replicates and interday analyses in 6 nonconsecutive working sessions (one per month).

Two different concentrations (20 and 3% v/v for C5K and 5 and 40% v/v for HFO) were assayed for each gas in the precision study. The results are expressed as the relative standard deviation (RSD %) in Table 4. RSD% values lower than 7.5% were obtained for both concentrations, which show the precision of the analytical method developed.

Table 4. RSD% Values Obtained for the Intermediate Precision for HFO and C5K.

2.2.4.

Open in a new tab

2.2.5. Identification of Decomposition Products Formed in the Medium-Voltage Electrical Switchgear

Different gas mixtures of HFO and C5K in combination with synthetic air or CO2 were selected as alternative gas mixtures to fill the cubicles of the four medium-voltage switchgears.

A permanent 30 kV AC voltage was applied to the four prototype MV switchgear cubicles to check the stability of these four gas mixtures under operating conditions in the electrical switchgear. For this purpose, the Ormazabal Corporate Technology’s experimentation network called UDEX (Demonstration and Experimentation Unit) has been used.16,17 This network is a highly configurable medium-voltage network independent from the grid, which allows the development and testing of new technologies, products, and services in a safe and controlled environment.

Both developed analytical methods were used for the quantification of C5K and HFO and for the detection, identification and, in some cases, quantification of decomposition products generated in the gas mixtures in the cubicles after being under permanent 30 kV AC voltage during a long time interval (4 years in cubicle #1 and 1.5 years in cubicles #2, #3, and #4).

In Figure 3, chromatograms of the gas samples of the identified compounds are shown. The initial compositions of HFO and C5K in gas mixtures used in the four medium-voltage cubicles and identified decomposition products are indicated in Table 5.

Figure 3.

Figure 3

Open in a new tab

Chromatograms of samples analyzed (a) by GC/MS and (b) by GC/TCD.

Table 5. Decomposition Compounds Identified in the Samples Studied.

2.2.5.

Open in a new tab

CF4 (tetrafluoromethane), C2F6 (hexafluoroethane), C3F6 (hexafluoropropylene), and C3HF7 (1,1,1,2,2,3,3-heptafluoropropane) are the degradation products from the C5K.

CF4 may be narcotic at high concentrations, thermally stable, and chemically very inert.6,18−20 C2F6 is relatively inert, nonflammable, and nontoxic, although asphyxiation may occur because of the displacement of oxygen.6,18,21,22 C3F6 is noncombustible, and it can asphyxiate by the displacement of air, may cause respiratory irritation, moderately or very reactive, and may be incompatible with many amines, nitrides, azo/diazo compounds, alkali metals, and epoxides.6,18,23 C3HF7 is heavier than air, and it may asphyxiate by the displacement of air, especially in a confined space. It does not react rapidly with air and with water. It may be incompatible with many amines, nitrides, azo/diazo compounds, alkali metals, and epoxides.6,18,21,22 CF4 and C3F6 appear in bibliography as moderate or slightly toxic decomposition products.3,23

CH2F2 (difluoromethane) and cis-C3H2F4 (1,3,3,3-tetrafluoropropene) are degradation products from hydrofluoroolefin. CH2F2 has high thermal stability. Its vapors are heavier than air. Under prolonged exposure to fire or intense heat, the containers may rupture violently and rocket.6,24 Cis and trans C3H2F4 are less flammable than the mixture of both cis and trans isomers.25,26

The physicochemical properties of these degradation compounds are summarized in Table 6, with the retention times and chemical formula.

Table 6. Retention Times (tr) and Physicochemical Properties of Identified Decomposition Compounds.

2.2.5.

Open in a new tab
*

Value referred to 15 months.

In view of the decomposition products that are formed and their abundance, the breakage of the C–C bond between C1 and C2 atoms of HFO and C5K (Table 1) is more likely to occur. This reaction generates CH2F2 for HFO and C3F6 y C3HF7 for C5K.

In the case of C5K, the C3F6 y C3HF7 compounds react further to produce CF4 and C2F6. Studies carried out by different authors27−30 corroborate the formation of the degradation products identified in our experiments for C5K. Degradation pathways according to the density functional theory (DFT) and modeling studies have been proposed in the literature for C5K. Furthermore, Wang, Y. et al.27 reported that C5K generates less CO with air mixtures and more oxygenated chemicals. However, under the conditions used in this work, CO or any other oxygenated compound is not produced in mixtures with air. CO is only generated in the CO2 medium. In addition, Zhang, Y. et al.30 reported that in the presence of O2, the formation of large molecules is inhibited and compounds with double bonds, such as C3F6, are not stable. Nevertheless, in this study, this compound is one of the major products. Therefore, the formation of decomposition products depends on the conditions of temperature, pressure, and voltage to which the gas mixtures are subjected.

The potential for all these degradation products of C5K and HFO to cause cancer in humans has not been assessed by the U.S. EPA IRIS program, the International Agency for Research on Cancer, or the U.S. National Toxicology Program 14th Report on Carcinogens (SRC).31,32 On the other hand, highly fluorinated compounds are not expected to biodegrade rapidly.33,34

The possible causes of these degradation products could be due to the influence of electrical field present in the cubicles and the interaction of compounds in the gas mixtures with water coming from HFO and from some plastic materials present inside the cubicles.

To evaluate if these decomposition products are acceptable for the adequate performance of the cubicles in an installation under real operating conditions over the entire life duration (30 years) is a more complex issue and requires more long-life assessments to check the behavior over time. New samples are expected to be extracted after an additional long period of time to follow the evolution of gas mixture compositions. Dielectric and mechanical type tests will have to be performed to the cubicles after a longer time to evaluate the influence of these new gas mixtures on the technical functionality of the medium-voltage switchgear cubicles.

After the identification of the degradation compounds produced in the different mixtures, the concentrations of the most abundant compounds (C3F6 and C3HF7) were obtained by GC/MS using a commercial gas mixture reference standard of 500 ppmv of each gas (C3HF7–CO–C3F6–He), Table 8.

Table 8. Variation of Response with the Concentration of C3F6 and C3HF7 by GC/MS.

2.2.5.

Open in a new tab

Furthermore, CO and water contents were analyzed in gas samples by GC/TCD and GC/MS, respectively, Table 7, using H2O (50 ppmv) and CO (500 ppmv) reference standards. H2O is analyzed because of its possible problem of generating degradation compounds and CO due to its toxicity.

Table 7. Concentration of CO, H2O, C3F6, and C3HF7 in the Samples Studied.

2.2.5.

Open in a new tab

Calibration curves, Table 8, for the acquired gas mixture reference standard of C3F6 and C3HF7 were built from 0.01 to 0.05% v/v, using the splitless mode in the GC/MS equipment and injecting a higher sample volume (0.5 mL).

Small quantities of weakly toxic C3F6 (≤0.04%), C3HF7 (≤0.4%), and a certain amount of CO (≤1%), H2O (≤0.006%) have been found in the mixtures.

2.3. Interlaboratory Studies

For further validation of the analytical method, an interlaboratory assay was carried out for C5K and the most abundant degradation compounds (C3F6 and C3HF7). Gas samples of three of the four cubicles (#1, #3, and #4) were extracted 3 months later than the gas samples analyzed in our laboratory. They were collected in this case in 500 cm3 stainless-steel cylinders to be sent to 3M Analytical Laboratory in USA.

3M Analytical Laboratory used GC-FID as the analytical technique and a semiquantitative method to calculate the C3F6 and C3HF7 contents in gas samples. C3F6 and C3HF7 concentrations were calculated by dividing their GC-FID chromatographic signal area by the chromatographic signal area of C5K, and the result was multiplied by the known concentration of C5K in the gas sample. The calculated values are shown in Table 9.

Table 9. Results of the Interlaboratory Assay Carried Out by Basque Country University (UPV/EHU) and 3M Analytical Laboratory.

2.3.

Open in a new tab

The results obtained in 3M laboratory appear to be very consistent with those obtained using the analytical method developed by our research group.

Additionally, the concentrations of the gas mixtures after these periods of electrical aging (4 years and 1.5 years) are very similar to the initial ones, and the four cubicles withstand 30 kV perfectly. In addition, the pressure of the gas mixtures in the cubicles is maintained (there are no abnormal leaks).

3. Conclusions

The analytical methodology for monitoring the chemical stability of dielectric gas mixtures at 30 kV in medium-voltage switchgear has been successfully developed and validated. In addition, the interlaboratory assay carried out has allowed us to obtain no significant differences in the results obtained by the two laboratories: UPV/EHU and 3M laboratory.

With the methodology validated, the decomposition products of gas mixtures of C5K and HFO with synthetic air and CO2 have been identified. CF4 (tetrafluoromethane), C2F6 (hexafluoroethane), C3F6 (hexafluoropropylene), and C3HF7 (1,1,1,2,2,3,3-heptafluoropropane) are the degradation products from C5K, and CH2F2 (Difluoromethane) and trans-C3H2F4 (1,3,3,3-Tetrafluoropropene) are the degradation products from HFO.

In this work, small quantities of weakly toxic C3F6 (≤0.04%), C3HF7 (≤0.4%), and a certain amount of CO (≤1%) and H2O (≤0.006%) have been found.

It is possible that part of C3HF7 dehydrofluorinates to form C3F6. The presence of C3HF7 in samples of gas mixtures with C5K can be explained from a reaction of fluoroketone C5K with the humidity present in plastics of the cubicle (for example, some polyamides).

One step to prevent the formation of these degradation products would be to avoid the presence of water within the cubicle. On the one hand, HFO contains humidity. On the other hand, the plastic components of the cubicles are polyamides, which contain 4% of water. Anyway, taking into account the values of GWP and LC50 (Table 6), the authors consider that the concentration levels of degradation products are acceptable in terms of toxicity. At any rate, using adequate adsorbing or absorbing materials, such as molecular sieve, could eliminate water from the gas mixture rapidly to avoid its reaction with C5K. However, the total reduction of water content is not possible because of the active surface properties of molecular sieve. Other alternative would be to use materials that do not retain humidity such as fluoro-elastomer rubber or butyl rubber.

All gas mixtures may be suitable as an alternative to SF6 because the low proportion of the decomposition products obtained do not represent a health or environmental problem. However, more long-life assessments to check the behavior over time of these decomposition products will be needed.

4. Material and Methods

4.1. Reagents

HFO was supplied by Inventec Performance Chemicals (Saint-Priest, France). Purity ≥99.5% weight. Moisture content ≤50 ppm weight and nonvolatile residue ≤50 ppm weight. C5K was obtained from 3M (Minnesota, USA). Purity 99.5%. CO2 and synthetic air were supplied by Carburos Metalicos and Air Liquid (Madrid, Spain) and were used as dilution gases for preparing the calibration curves of the different dielectric gases assayed. The decomposition products CO, C3F6, and C3HF7 were acquired as a mixture of 500 ppmv of each gas in He (CO–C3HF7–C3F6–He) from Air Liquid (Madrid, Spain) to be used as a gas reference standard. Finally, an additional gas reference standard with 50 ppmv of water (H2O/N2.B10.IMP) has been obtained from Air Liquid (Madrid, Spain). Degradation products are of crystal quality, recommended for calibrating analyzers and adjustment equipment for instrumentation, and with a manufacturing tolerance between 5 and 10% and an uncertainty between 2 and 5%.

4.2. Instrumentation

GC analysis was performed using a gas chromatograph Agilent 7820A (California, USA), equipped with MS and TCD detectors (5975C and G4332A, respectively).

4.3. Chromatographic Conditions

A GC equipment with two injectors was used in this work. One is connected to a column Poraplot Q (25 m × 0.25 mm × 8 μm) and an MS detector. The other injector is joined to two columns in serial connection, a capillary column HP-PlotQ (30 m × 0.53 mm × 40 μm) and the other one is a HP-Molesieve (30 m × 0.32 mm × 12 μm). These are connected by a system of one valve that opens and closes to allow the gas sample to pass from one to another, intermittently, until finally communicating with the TCD detector.

The mechanism of this type of column is not only based on the difference in the boiling point/vapor pressure or in polarities but also by the different interactions of the gas with the stationary phase. PoraPLOT Q is a fused silica recommended column for systems that analyze polar and apolar volatile compounds. HP-PLOT Q is a bonded polystyrene-divinylbenzene column. It was specially developed for the separation of targeted nonpolar and polar compounds, including hydrocarbons, CO2, methane, air, CO, water, polar solvents, and sulfur compounds (H2S). HP-Molesieve columns, made of zeolite, are ideal for many applications, including routine air monitoring for the analysis of fixed and permanent gases: H2, O2, N2, CO, CH4, He, Ne, Ar, Kr, Xe, and SF6. Therefore, the combination of these columns, together with the detectors, is appropriate to be able to identify any decomposition product of this type of fluorinated gases.

In order to obtain a good sensitivity of the method and a good resolution of chromatographic peaks, the syringe to inject gas without leaks, the liner, the volume to be injected, the split ratio in each system, injector conditions, and the separation temperature ramp were optimized.

The sample was injected into a chromatograph using a Hamilton syringe glue-free HDHT (GR, Switzerland) after sampling from the working Tedlar bags. Volumes of the inserted sample were 0.1 mL at 1:10 split ratio for TCD and 1 mL at 1:55 split ratio for MS, using a liner 5183–4647 (Agilent, California, USA). The temperature of MS and TCD injectors was set at 200 °C and pressure at 2.4 psi and 14 psi, respectively. The oven temperature was programmed with an initial temperature of 40 °C for 5 min, followed by an increase at a rate of 10 °C min–1 to 75 °C; finally, the oven temperature was increased at a rate of 5 °C min–1 up to 200 °C.

The analysis of the mass spectra obtained allows identifying the decomposition products by chemical interpretation or comparison with the NIST14 mass spectra database. Mass spectra were collected from m/z 15 to 300 (with an electronic impact ionization source of 70 eV).

4.4. Sampling

The four medium-voltage switchgear cubicles used in this experimentation (Figure 1) are manufactured according to international standard IEC 62271–200:2011 “High-voltage switchgear and controlgear-Part 200: AC metal-enclosed switchgear and controlgear for rated voltages above 1 kV and up to and including 52 kV” (https://webstore.iec.ch/publication/6716) and IEC 62271–1:2017 “High-voltage switchgear and controlgear-Part 1: Common specifications for alternating current switchgear and controlgear” (https://webstore.iec.ch/publication/32982), their rated voltage being 36 kV and their rated current 630A. They all have a hermetic tank made of AISI 304 stainless steel that contains the medium-voltage circuit and that is filled with each gas mixture. Each manufacturer of medium-voltage switchgear has its own designs and models, with some differences in the construction details. The dimensions of the gas tanks where the gas mixtures are contained are shown in Figure 4.

Figure 4.

Figure 4

Open in a new tab

Dimensions of the medium-voltage switchgear cubicles used.

These four medium-voltage switchgear cubicles were filled at 20 °C with four different alternative gas mixtures, with a final absolute pressure of 1400 mbar (Table 10), and subjected to permanent 30 kV AC voltage. The electric field in the gas tanks is obviously nonhomogeneous, but in critical areas inside (that is, near to epoxy resin busbars), we have calculated that a maximum electric field is 3.41 kV/mm. The dielectric strength of gas mixtures was determined according to the method ASTM D2477 (standard test method for dielectric breakdown voltage and dielectric strength of insulating gases at commercial power frequencies). The four cubicles filled with the four different gas mixtures passed a standardized dielectric test (70 kV AC withstand the voltage test at a power frequency of 50 Hz) without dielectric failure. The cubicles have not been operated during aging, 4 years (cubicle #1) and 1.5 years (cubicles #2, #3, and #4). After this electrical aging time, gas samples were extracted from the cubicles and taken in 500 cm3 Tedlar bags, Vertex (Barcelona, Spain), to be analyzed on the same day by GC/MS and GC/TCD.

Table 10. Gas Mixture Samples Analyzed.

4.4.

Open in a new tab

4.5. Optimization and Validation of GC–MS and GC-TCD Methods

Optimization of the chromatographic method was carried out. The most suitable conditions for the injection and elution were studied in order to obtain the maximum sensitivity and repeatability, along with an acceptable chromatographic separation in terms of resolution. The optimized method was validated by evaluating the following parameters: selectivity, LOQ, linear concentration range, and intermediate precision. Finally, for further validation of the analytical method, an interlaboratory assay was carried out in collaboration with 3M Analytical Laboratory (Minnesota, USA).

Acknowledgments

Authors thank 3M Company for interlaboratory assay.

This work was supported the Basque Country Government (project Elkartek KK-2017/00090).

The authors declare no competing financial interest.

References

  1. https://www.ipcc.ch [accessed 1 May 2020].
  2. Tsai W. The decomposition products of sulfur hexafluoride (SF6): reviews of environmental and health risk analysis. J. Fluorine Chem. 2007, 128, 1345–1352. 10.1016/j.jfluchem.2007.06.008. [DOI] [Google Scholar]
  3. Tatarinov A. V.; Bilera I. V.; Avtaeva S. V.; Shakhatov V. A.; Solomakhin P. V.; Maladen R.; Prévé C.; Piccoz D. Dielectric barrier discharge processing of trans-CF3CH=CHF and CF3C(O)CF(CF3)2, their mixtures with air, N2, CO2 and analysis of their decomposition products. Plasma Chem. Plasma Process. 2015, 35, 845–862. 10.1007/s11090-015-9635-8. [DOI] [Google Scholar]
  4. Seeger M.; Smeets R.; Yan J.; Itod H.; Claessens M.; Dullni E.; Falkingham L.; Franck C. M.; Gentils F.; Hartmann W.; Kieffel Y.; Jia S.; Jones G.; Mantilla J.; Pawar S.; Rabie M.; Robin-Jouan P.; Schelleken H.; Spencer J.; Uchii T.; Li X.; Yanabu S. Recent trends in development of high voltage circuit breakers with SF6 alternative gases. PPT 2017, 4, 8–12. 10.14311/ppt.2017.1.8. [DOI] [Google Scholar]
  5. Beroual A.; Haddad A. Recent Advances in the Quest for a New Insulation Gas with a Low Impact on the Environment to Replace Sulfur Hexafluoride (SF6) Gas in High-Voltage Power Network Applications. Energies 2017, 10, 1–20. 10.3390/en10081216. [DOI] [Google Scholar]
  6. https://pubchem.ncbi.nlm.nih.gov [accessed 1 March 2021].
  7. Wang Y.; Gao Z.; Wang B.; Zhou W.; Yu P.; Luo Y. Synthesis and Dielectric Properties of Trifluoromethanesulfonyl Fluoride: An Alternative Gas to SF6. Ind. Eng. Chem. Res. 2019, 58, 21913–21920. 10.1021/acs.iecr.9b05633. [DOI] [Google Scholar]
  8. Zhang X.; Li Y.; Chen D.; Xiao S.; Tian S.; Tang J.; Wang D. Dissociative adsorption of environment-friendly insulating medium C3F7CN on Cu(111) and Al(111) surface: A theoretical evaluation. Appl. Surf. Sci. 2018, 434, 549–560. 10.1016/j.apsusc.2017.10.201. [DOI] [Google Scholar]
  9. Preve C.; Piccoz D.; Maladen R.. Presented in part at 23rd International Conference on Electricity Distribution; Lyon: June, 2015. [Google Scholar]
  10. Safety data sheet of Solstice-1234ze, according to Regulation (EC)No.1907/2006. http://www.harpintl.com/downloads/pdf/msds/Solstice-1234ze-SDS-CLP.pdf [accessed 1 March 2021].
  11. Tatarinov A. V.; Bilera I. V.; Shakhatov V. A.; Avtaeva S. V.; Solomakhin P. V.; Maladen R.; Prévé C.; Piccoz D. Comparative Study of Degradation of trans-1,3,3,3- Trifluoropropene, 2,3,3,3-Tetrafluoropropene, Perfluoro-3-methylbutanone-2, and Sulfur Hexafluoride in Dielectric Barrier Discharge. High Energy Chem. 2016, 50, 64–70. 10.1134/S0018143916010100. [DOI] [Google Scholar]
  12. Kessler F.; Sarfert-Gast W.; Ise M.; Goll F.. Presented in part at the 20th International Symposium on High Voltage Engineering; Argentina, August 2017. [Google Scholar]
  13. Lutz B.; Juhre K.; Kuschel M.; Glaubitz P. CIGRÉ Colloquium Study Committees, http://www.cigre.org, 2017
  14. Zhang X.; Tian S.; Xiao S.; Deng Z.; Tang J. Insulation Strength and Decomposition Characteristics of a C6F12O and N2 Gas Mixture. Energies 2017, 10, 1170. 10.3390/en10081170. [DOI] [Google Scholar]
  15. Lokhat D.; Ramjugernath D.; Starzak M. Simultaneous Separation of perfluorocarbons and Oxyfluorocarbons Using Capillary Gas Chromatography:Analysis of the Products of Hexafluoropropene Oxidation. J. Anal. Chem. 2014, 69, 98–103. 10.1134/S1061934814010109. [DOI] [Google Scholar]
  16. Orue I.; Gilbert I.; Larrieta J.; Sánchez J.A.. Presented in part at the CIRED 24th International conference on Electricity Distribution; Glasgow, June 2017. [Google Scholar]
  17. Barrios S.; Gilbert I.; Hurtado A.; Muklroy P.; Orue I.. Presented in Smargrinet; Croatia, April 2018. [Google Scholar]
  18. https://www.cdc.gov/niosh/index.htm. [accessed 1 March 2021].
  19. Pohanish R.P.Sittig’s Handbook of Toxic and Hazardous Chemical Carcinogens; 6th ed. New York: Waltham; 2012. [Google Scholar]
  20. O’Neil M.J.The Merck Index - An Encyclopedia of Chemicals, Drugs, and Biologicals; 15th ed.; Royal Society of Chemistry: Cambridge, 2013; 1–2708. [Google Scholar]
  21. Lewis R.J.Hawley’s Condensed Chemical Dictionary; 15th ed.; John Wiley & Sons: New York,2007; 1–4000. [Google Scholar]
  22. U.S. National Archives and Records Administration’s Electronic Code of Federal Regulations. http://www.ecfr.gov. [accessed 1 March 2021].
  23. Bretherick L. Handbook of Reactive Chemical Hazards; 4th ed.; Butterworth-Heinemann Ltd.: Boston, 1990;1–2058. [Google Scholar]
  24. https://unfccc.int. [accessed 1 March 2021].
  25. Gong M.; Zhao Y.; Dong X. Measurements of isothermal (vapor + liquid) equilibrium for the (propane + cis-1,3,3,3-tetrafluoropropene) system at temperatures from (253.150 to 293.150) K. The Journal of Chemical Thermodynamics 2016, 103, 349–354. 10.1016/j.jct.2016.08.031. [DOI] [Google Scholar]
  26. Liang Y.; Wang F.; Sun Q.; Zhai Y. Molecular structural and electrical properties of trans-1,3,3,3-tetrafluoropropene and 2,3,3,3-tetrafluoropropene under external electric fields. Comput. Theor. Chem. 2017, 1120, 79–83. 10.1016/j.comptc.2017.09.017. [DOI] [Google Scholar]
  27. Wang Y.; Huang D.; Liu J.; Zhang Y.; Zeng L. Alternative Environmentally Friendly Insulating Gases for SF6. Processes 2019, 7, 216–230. 10.3390/pr7040216. [DOI] [Google Scholar]
  28. Zhang X.; Li Y.; Tian S.; Xiao S.; Chen D.; Tang J.; Zhuo R. Decomposition mechanism of the C5-PFK/CO2 gas mixture as an alternative gas for SF6. Chem. Eng. J. 2018, 336, 38–46. 10.1016/j.cej.2017.11.051. [DOI] [Google Scholar]
  29. Zhang X.; Li Y.; Xiao S.; Tang J.; Tian S.; Deng Z. Decomposition Mechanism of C5F10O: An Environmentally Friendly Insulation Medium. Environ. Sci. Technol. 2017, 51, 10127. 10.1021/acs.est.7b02419. [DOI] [PubMed] [Google Scholar]
  30. Zhang Y.; Zhang X.; Li Y.; Li Y.; Chen Q.; Zhang G.; Xiao S.; Tang J. AC breakdown and decomposition characteristics of environmental friendly gas C5F10O/Air and C5F10O/N2. IEEE Access 2019, 7, 73954–73960. 10.1109/ACCESS.2019.2915372. [DOI] [Google Scholar]
  31. https://www.epa.gov⟩ iris. [accessed 1 March 2021].
  32. https://www.iarc.fr. [accessed 1 March 2021].
  33. https://ntp.niehs.nih.gov⟩ roc14. [accessed 1 March 2021].
  34. Boethling R. S.; Meylan P. H.; Meylan W.; Stiteler W.; Beauman J.; Tirado N. Group contribution method for predicting probability and rate of aerobic biodegradation. Environ. Sci. Technol. 1994, 28, 459–465. 10.1021/es00052a018. [DOI] [PubMed] [Google Scholar]

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES