1. Introduction
In northwest of China, there are many UHV transmission lines in the higher altitude regions, where strong sandstorm weather accounts for a considerable proportion in the whole year. Corona caused by this special weather is one of the key factors for the selection and optimization of UHV transmission lines. It is important to study the corona characteristic of the conductors in the sandy and dusty condition in high-altitude areas [1-5].
A few of such experimental studies have been done. In 1941, Bagnold first [6] introduced tunnel to physical study for sandy and windy flow. Arainy et al used a test system for dusty weather simulation to research air-gap breakdown under dusty circumstance [7-11]. Mei et al developed a set of portable tunnels to study the wind erosion research [12]. Lin et al used a test system for dusty weather simulation to research the contamination deposited characteristics of porcelain long rod insulators under sandstorm circumstance [13]. In these studies, the structures of the sand-wind tunnel are closed and the variants are fixed [14-15]. However, there are many shortcomings of those devices, such as lacking of regulation precision, inconvenient to use in different places because of huge volume and so on [16-19]. In order to consider the limitations of the off-site test and site constraints at different altitudes, a set of movable sandy and dusty condition simulation system for high altitude test based on corona cage is to be designed. The wind speed and the dust concentration of the system could be controlled in the range of experiment, the test precision could meet the corresponding require-ments. The system should have the advantages of small occupation area and easy disassemly and transportation The following requirements of experimental design are proposed according to the results of the numerical simulation test system:
1) Based on the comprehensive consideration of actual situation in northwest region and the feasibility of the test system, the range of wind speed is set between 5 m/s and 15 m/s. The sandy and dusty concentration (sand ratio) is set in the magnitude of 10−8~10−7. Wind speed, sandy and dusty concentration are adjustable with the accuracy according to the requirement of the test. 2) Sandy and dusty coverage needs cover the range of motion of ion which is corona discharge. Through calculation by charge-simulation method, the sandy and dusty coverage is finalized as: it should be longer than 6m in the direction along the conductors; it should be longer than 2m in the vertical along the conductors, as shown in Fig. 1. 3) In the northwest region, the parameters of sand and dust for numerical simulation and experimental study are selected as follow: particle size is set as 50~200μm; the density of particle is set as 2200kg·m−3. 4) The test will be performed in different place at different altitude. With the consideration of the condition of transportation and test site, some requirements for the test system are simple structure, easy to assemble, disassemble, transport, operation, and in small space occupied under the premise of meeting the design parameters.
Fig. 1.Schematic diagram of dusty coverage
2. Design of Structure of the Experimental System
2.1 The proposed preliminary program
This paper proposed four programs with rectangular or circular duct. And the programs are shown in Fig. 2
Fig. 2.Schematic diagrams of each program
In order to facilitate the programs shown in Fig. 2 for comparative analysis, the concept of “particle diffusion tube length” is defined, namely: particle diffusion tube length (Lk) which is shown in Fig. 3 is the length from the outlet of feeding tube to the place where sand initially fill the duct in a certain wind speed.
Fig. 3.Schematic diagram of particle diffusion tube length with wind speed in rectangular air-duct
2.2 Numerical simulation and comparative analysis of each program
In the commercial software of Gambit, the works of establishment of model, meshing and definition of boundary condition is done. The schematic diagram of grids of program 4 is shown in Fig. 4, as an example. By the independent verification, the number of grids of this program is ultimately determined as about 56 thousand. The inlet of air-duct and the outlet of feeding pipe are defined as the boundary condition of velocity inlet, and the outlet of air-duct is defined as the boundary condition of out flow.
Fig. 4.Schematic diagram of grid of program 4
The model is introduced into commercial software of Fluent for numerical simulation. The assumption of the modle of standard k- turbulence is that the flow which will to be simulated is fully turbulent. Taking that the flow is fully turbulent into account, the model of standard k− ε turbulence is selected to simulate the flow field [20].
The continuity equation:
The momentum equation :
where is time-averaged velocity; is time-averaged pressure; i and j is the coordinate direction in the 3D Cartesian coordinate system, i, j =1, 2, 3; μ and μT respectively are the viscosity ratio of motion of the fluid and viscosity ratio of the turbulent.
The turbulent kinetic energy k equation and the dissipation rate ε equation of standard k−ε model:
where Gk is the term of turbulent kinetic energy k which is caused by average pressure gradient; Gb is the term of turbulent kinetic energy k which is caused by the effect of buoyancy; YM is the impact of compressible turbulent pulsation to the total dissipation rate; C1ε, C2ε and C3ε is empirical constants; σk and σε respectively are the number of Prandtl which correspond with the turbulent kinetic energy k and the dissipation rate ε.
In this paper, the sand ratio of the sandy and dusty waether which is to be simulated is at the order of magnitude of 10−8~10−7. In the other words, the volume fraction of discrete phase is very low. Therefor, the DPM model which is suitable for the condition of low volume fraction is selected [21-23], whose equations (coordinate direction of X in the 3D Cartesian coordinate system) of particle motion of discrete phase is:
where va and vp respectively are the velocity of gap-phase and particle, m/s; ρa and ρp respectively are the density of gap-phase and particle, kg/m3 ; t is the movement time, s; gx is the acceleration of gravity in x direction, m/s2 ; fx is the sun of all kinds of additional force on unit mass of particle, N; the second term at the right of the equal sign is the sun of gravity and buoyancy on unit mass of particle; fD(va-vp) is the drag force of unit mass of particle, and
where μ is gas-phase viscosity coefficient, Pa·s: ρp is the particle density, kg/m3 ; dp is the particle diameter, m; CD is the drag coefficient; Rep is the Reynolds number of particle.
In this simulation, the refrigerant of gas is air. Comprehensive consideration the maximum wind speed of each month in many years in the high-altitude areas and the feasibility of the test, the wind speed of this test is selected as 5~15m/s. According to the properties of the sand in northwest of China (showed in Table 1), the particle size and density of the sand are selected respectively as 50~200μm and 2200kg/m3 .
Table 1.The properties of the sand in northwest of China
For the programs with rectangular air-duct, according to the demand of parameters of fans and wind speed, the structure of outlet of the duct is selected as 3×1.5m2. Fig. 5 shows the results of numerical simulation of particle diffusion tube length in rectangular air-duct as function of wind speed. Fig. 6 shows the length of sandy and dusty coverage as function of wind speed in rectangular air-duct. Fig. 7 shows the diffusion of sand and dust in the corona cage at different wind speeds in program 2.
Fig. 5.Change of particle diffusion tube length
Fig. 6.Change of maximum length of sandy and dusty coverage with wind speed in rectangular air-duct
Fig. 7.Diffusion of sand and dust in the corona cage at different wind speed in program two
Fig. 5 shows that duct particle diffusion tube length increases with wind speed increases in the rectangular straight. This system can not satisfy the miniaturization requirements of small footprint, and the proliferation of symmetry is poor. Fig. 6 and Fig. 7 shows that, although in the program two, dust cover meets the requirements of the length not less than 6m, but its export diversion structure make the continuity of sand-dust diffusion in the corona cage poor, and the length of sandy and dusty coverage changes. In the test, in order to ensure the length of sandy and dusty coverage is enough in corona cage, the length of air-duct need to be changed to accommodate it, which makes the operation of the test is complicated and difficult; the control of wind speed, sandy and dusty concentration is impacted. Therefore, programs two can not meet the requirement of the test
For the programs with round air-duct, according to the demand of parameters of fans and wind speed, the diameter of air-duct is selected as 1200mm. To ensure the sandy and dusty coverage, three sets of same test systems which are arranged side by side are pre-set. Therefore, the height and length of sandy and dusty coverage is more than 2m in every single of the test system.
In order to accelerate the diffusion of dispersed phase and to make the length of tube shorter, axial guiding vanes are installed in front of the feeder opening to generate swirl wind. The numerical simulation results show that when the wind speed is set as 5~15m·s−1, particle diffusion tube length can be controlled within 3m by adjusting the guiding deflection angle of vanes (0~45°). It is known as the practical technology of particle diffusion length controlled by swirl, as shown in Fig. 8.
Fig. 8.Renderings of simulation of swirl dispersion
Fig. 9 shows the results of numerical simulation of the length of sandy and dusty coverage in corona cage when the test is done with programs with circular air-duct adopting practical technology of particle diffusion length controlled by swirl in which effective sandy and dusty coverage starts from the place of 3m from the exit of air-duct.
Fig. 9.Change of length of sandy and dusty coverage with wind speed in round air-duct
Fig. 9 shows that both programs with circular air-duct can meet the need that sandy and dusty coverage is more than 2m; but according to the position of conductors in corona cage, fans need to be arranged overhead, so the problem of vibration of fans may seriously affect the security of test system, in program three. On the other hand, program 4 combines the advantages that length of sandy and dusty coverage is enough in straight pipe and fans can be arranged on the ground to reduce the influence of vibration of fans to the security of the test system. The covering of the test system is 9.5×2m2 in this program. The requirement of small space occupied is met. Therefore, program 4 is selected as final program
In order to further improve the design of the test system and to make it easy to disassemble and transport, three-section design and bolted connection are adopted in the support structure when the test is done in different places. The schematic diagram is shown in Fig. 10.
Fig. 10.Schematic diagram of supporting structure in program 4
2.3 Design of control of the test system
A single set of the test system uses two parallel axial fans. In order to obtain uniform concentration of sandwind flow, screw feeder, which is based on the principle of volume pump, is selected. The regulation characteristic of the feeder is that feeding quantity changes linearly with rotation speed. When rotation speed is unchanged, the capacity of feeding is unchanged. In order to obtain the expected velocity and sandy and dusty concentration, the inverter which has the advantage of higher accuracy of regulation and easy operation is adopted to regulate the speed of fans and spiral feeder. The technical parameters of main equipment of the test system and the process of regulation and control are shown in Table 2 and Fig. 11 respectively. Because the fans in a single set of the test system are parallel, the inverter adopts the mode of one for two.
Table 2.Parameters of test system main equipment
Fig. 11.Logic diagram of frequency control
2.4 Design of measurement of the test system
To reduce the impact of complex flow field of short tube to the measurement of flow, in this paper, the method of flow measuring in double section is adopted to measure the flow. Namely, Annubar flowmeter whose structure and parameters are as shown in Fig. 12 and Table 3 is adopted to obtain the instantaneous flow value of air in the end of high-horizontal section. The law of Backrest tube traverse is adopted to obtain real time mean flow values. Mixed flow values are obtained after further processing. In order to understand the specific sand concentration at the position of conductor, environmentdust monitoring system which is based on the electrostatic induction technique is adopted (as shown in Table 4 [24]. The system is composed by dust monitoring (DUMO), module of data acquisition (EDA9017), converter (RS485-USB) and software of data acquisition, where DUMO is base on induction technology of electrostatic. Light, movable and height adjustable support platform is designed to make it easy to install on the corona cage.
Fig. 12.Annubar flowmeter: (a) Piezometric tube; (b) Three groups of valve and differential pressure transmitter
Table 3.Parameters of Annubar flowmeter
Table 4.Parameters of environmental dust monitoring system
3. Test System Performance Tuning
Fig. 13 shows the actual figure of the test system. Firstly, tuning of frequency and flow performance is done at different opening of deflectors at head and hail of high level segment. The performance of flow is obtained at the opening of 25-30 which is the best opening for sand diffusion, as is shown in Fig. 14. With comprehensive consideration of characteristics of flow at every opening, the range of wind speed of the system is set as 4-16m·s−1.
Fig. 13.Achieve rendering of the test system
Fig. 14.Characteristic of flow at opening 25-30
After the local sand material in Xining is screened, the sand, the size of which is smaller than 0.25mm, is chosen to measure concentration when the opening is 25-30. The results are shown in Fig. 15, where fg is the power frequency of feeder. It can be seen from Fig. 13 that the sandy and dusty density of the test system at the conductors position is equal or greater than 133mg·m−3 (sand ratio 6×10−8).
Fig. 15.Characteristic of dusty concentration at opening 25-30
Tuning results show that the test system can meet the requirements of concentration, wind speed and dusty coverage.
4. Analysis of Preliminary Test Results of Corona Characteristic
Fig. 16 shows the influence of different sandy and dusty density and different altitudes on corona characteristic of conductor. As the density of sandy and dusty increases in the same field strength, corona loss increases significantly while the corona onset voltage decreases significantly. The test system is well designed to meet the test requirements of corona characteristic of conductors.
Fig. 16.Influences of different sand and dusty density and different altitude on corona characteristics of conductors
The researches about experimental corona characteristics at different altitudes, and “current-voltage characteristics” of the corona discharge have been completed by other numbers of our research group on the basis of the test system which is designed in this paper. And some relevant papers are published. In these papers: literature [25] used this experimental system to in-depth study the impact of the 0-3000m range of altitude on the corona onset characteristics of AC transmission line conductor; literature [26] using this experimental system to study the corona onset voltage of six split conductors at 19-4000 m range of altitude, and the “current-voltage characteristics” of the corona discharge of different corona cage split conductors is acquired. Because of the limited of space, this paper is not willing to further instruct the works about these aspects.
5. Conclusions
In general, this study has shown the following conclusions:
1) The wind speed range of the test system designed for 4~16m/s, the dust concentration larger or equal to 133 mg·m−3 (sand air ratio 6×10−8) and the dust cover of the wire is not less than the range of 6×2 m2 , the performance parameters meet the design requirements; 2) Defines the concept of particle diffusion tube length, cyclone obtained by numerical simulation of particle diffusion tube length control practical technology enables control of particle diffusion tube length 3m or less, a single system covering 9.2×2m2 , which meet the requirements for small area; 3) Testing system uses segmented structure and bolted to meet the needs of remote testing with easy disassembly and transport; 4) Using this test system can obtain different characteristics in different corona wire dust concentration under high altitude.
References
- L. Xing, Q. Sun, Y. P. Liu, et al. “Altitude correction methods for corona loss of 750 kV transmission line,” High Voltage Apparatus, vol. 46, no. 7, pp. 20-22, Jul. 2010.
- Y. P. Liu, S. H. You, F. C. Lv, et al, “Analysis of single stranded conductor’s negative DC corona onset characteristic at different altitudes,” High Voltage Engineering, vol. 36, no. 10, pp. 2417-2423, Oct. 2010.
- J. Tang, Y. P. Liu, X. Wu, et al, “Effect of altitude on radio interference based on corona test cage,” High Voltage Engineering, vol. 35, No. 3, pp. 601-606, Mar. 2009.
- F. C. Lv, S. H. You, Y. P. Liu, et al, “Calculation and analysis of bundle conductor’s AC corona loss in corona cage,” High Voltage Engineering, vol. 37, no.12, pp. 2878-2884, Dec. 2011.
- Y. P. Liu, Y. H. Li, et al. “Design of simulation test system for corona characteristics of conductor under sandstorm condition,” High Voltage Engineering, vol. 38, no. 9, pp. 2417-2423, Sep. 2012.
- Bagnold R A. “The physics of blown sand and desert dunes,” London: Mathuen&Co. Ltd, 1941.
- Qureshi M I, Al-Arainy A A, Malik N H, “Proformance of rod-rod gaps in the presence of dust particles under lightning impulses,” IEEE Transactionon Power Delivery, vol.6, no. 2, pp. 706-714, 1991. https://doi.org/10.1109/61.131130
- Al-Arainy A A, Malik N H. Qureshi M I, “Influence of desert pollution on the lightning impulse breakdown voltage of rod to plane air gap,” IEEE Transaction on Power Delivery, vol.6, no. 1, pp. 421-428, 1991. https://doi.org/10.1109/61.103767
- Qureshi M I, Al-Arainy A A, Malik N H, “Influence of sand/dust contamination on the breakdown of asymmetrical air gaps under lightning impulses,” IEEE Transaction on Electrical Insulation, vol. 7, no. 2, pp. 193-206, 1992.
- Qureshi M I, Al-Arainy A A, Malik N H, “Performance of rod-rod gaps in the presence of dust particles under standard switching impulses,” IEEE Transaction on Power Delivery, vol. 8, no. 3, pp. 1045-1051, 1993. https://doi.org/10.1109/61.252630
- Qureshi M I, Al-Arainy A A, Malik N H, “Influence of sand/dust contamination on the breakdown of asymmetrical air gaps under switching impulses,” IEEE Transaction on Dielectrics and Electrical Insulation, vol. 1, no. 2, pp. 305-314, 1984.
- M.Dong, “An Experimental study on the performance of sand sampler and sand discharger performance of the portable wind erosion-wind tunnel,” Inner Mongolia Agricultural University, 2005.
- L. Yang, Y. P. Hao, L. C. LI, et al, “Wind tunnel test for contamination deposited characteristics of porcelain long rod insulators under sandstorm circunstance,” High Voltage Engineering, vol, 36. no. 11, pp. 2651-2656, Nov.2010.
- F. Lu, “Power grid expressway shows its exhibition in northwest - the summarization of construction of 750kV backbone network in northwest by state grid,” China Power Enterprise Management, no. 25, pp. 59-60, Sep. 2009.
- H. L. Jia, L. Xu, S. X. Wei, et al, “The analysis of occurrence regulation of wind and sandstorm in Qinghai province,” Qinghai Science and Technology, no. 5, pp. 39-40, 2002.
- Y. H. Shi, “Research on the Mechanism of Discharge of Transmission Line in Sand-dust Storm,” Soochow University, 2011.
- Q. M. Sun, Z. Z. Li, S. L. Wu, et al, “The Summarization of Research Accomplishment in Sand Environment Wind Tunnel,” Journal of Xinjiang Normal University (Natural Sciences Edition), vol, 25, no. 3, pp. 106-111, Sep.2006.
- J. J. Qu, M. H. Yan, et al, “Wind tunnel simulation of sandstorm electrification,” Science China, vol. 33, no. 6, pp. 593-601, Jun. 2004.
- J. J. Qu, R. P. Zu, et al, “Wind tunnel simulation of effect of sandstorm on electrical wire voltage,” Journal of Desert Research, vol. 24, no. 5, pp. 534-538, Sep. 2004.
- S. Z. Wu, C. Y. Ren, “Numerical simulation of wind blown sand based on the Eulerian model,” Journal of Lanzhou University (Natural Sciences), vol. 48, no. 1, Feb. 2012.
- S. N. Hui, T. M. Xu, et al, “Study on the characteristics of gas-solid two-phase flow in a lain elbow,” Journal of Xi’An Jiaotong University, vol. 24, no. 4, pp. 65-73, Aug. 1990.
- G. S. Liu, S. Y. Chen, et al, “The experimental study on gas-solid two-phase flow in bend pipes of off-stoichiometric burner,” Journal of Engineering Thermophysics, vol. 15, no. 4, pp. 441-445, Nov. 1994.
- S. Z. Huang, Z. R. Yan, “Numerical simulation for two phase flow of air & solid in vertically up to horizontal and circular bend,” Boiler Technology, vol. 35, no. 5, pp. 9-11, Sep. 2004.
- “The instructions of environment-dust monitoring system Ver: 1.0,” Beijing Pakelaisi Analysis Instrument Co Ltd.
- F. C. Lv, Q. Z. Geng, L. Zhu, et al, “Research on corona inception characteristics of 750 kV transmission line conductors at different altitudes,” High Voltage Apparatus, vol. 49, no. 9, pp. 1-6, Sep. 2013.
- S. H.You, Y. P. Liu, F. C. L V, et al, "Calculation and analysis on corona onset voltage of corona cage bundle conductors at different altitudes," vol. 32, no. 4, pp.169-177, Feb. 2012.