Reliability Estimation of Electricity Distribution Substation Surge Protection System Composed by Surge Arresters with Different Operational Parameters Nikolay Nikolov1), Neli Dimitrova2), Anton Georgiev1) and Margreta Vasileva 2) 1) Department of Electronic Engineering and Microelectronics, 2) Department Electric Power Engineering, Technical University of Varna, Varna, Bulgaria
[email protected]
Abstract. The aim of this article is to suggest an approach to obtaining estimation values regarding reliability of electricity distribution substation surge protection system in cases when it is composed by surge arresters characterized by different operational parameters. A benchmarking is performed in order to select a set of surge arresters providing the surge protection system with a high reliability. Based on it some practical recommendations are given to be considered applying the approach suggested regarding surge protection system in operation. Keywords: reliability, surge protection, overvoltage.
1
Introduction
A very important part of the electricity distribution middle voltage and high voltage systems are the surge protection systems (SPS) [1]. These are intended to protect the substation equipment against overvoltages resulting from lightning strokes on the substation area and particularly on the incoming/outgoing overhead lines (OHL). The present study concerns only this part of a surge protection system which is composed by surge protection devices, such as surge arresters (SA). Within the study below this is called SPS. Most often the basic and auxiliary substation equipment is protected against overvoltages of atmospheric origin by metal-oxide SA and the SPS system is a set of such. A very important task standing in front of operational engineers is to perform the right selection of a proper set of SA so that to ensure as high as possible equipment protection level [5]. A key role in fulfilment of this play the overvoltages values expected, as well as the technical characteristics of SA [8]. The usual approach to selection of suitable SA set is to consider data collected during the substation operation and the operational experience of the engineers as a basis of this, taking also into account the referential data of the SA types at the disposal. Such approach does not always guarantee a high SPS reliability as this is not based on SPS reliability assessment. As it is known from the reliability theory [6] the first precondition for to provide a system with a high reliability is this to be composed by adfa, p. 1, 2011. © Springer-Verlag Berlin Heidelberg 2011
reliable components. For an SPS these are metal-oxide SA and their respective groundings. It is also important the system reliability to be assessed considering its components interconnection and interaction in terms of reliability. Following a classical approach to reliability assessment of an SPS there is developed and presented in [3] a mathematical model of substation SPS reliability there. This is chosen as a basis for the current study, which object is to determine a wellgrounded approach to a proper SA set selection, based on SPS reliability assessment. The SPS reliability model cited is derived based on assumptions in regard to the system elements as follows: their failures are random and independent events; their failure rate does not depend on time, hence the uptime distribution can be described by the exponential random variables distribution; they are characterized by rated values of their parameters and operate within the limits of their operational mode, specified in the technical documentation.
2
Reliability estimation of SPS
In order to make the study more perspicuous, a particular case is analyzed below. Consider an electricity distribution substation presented by its single-line diagram, shown on Fig.1. The substation is supplied by three 110 kV incoming overhead lines (OHL). It is composed in line of single bus system with bus sectionalizer and includes two 110/20 kV, 25 MVA power transformers.
Fig. 1. Single-line diagram of a 110/20 kV substation under analysis
Usually SA are installed in groups of three at each of OHL, connected respectively one by one between each of the phase lines and each of the groundings. Such groups of SA are installed also at the power supply side of the power transformers. Another pairs of one SA and one grounding are connected to the neutral at the power supply side of each of power transformers. Let denote the SA installation points as follows: at an incoming OHL as Position 1, at a power transformer as Position 2 and at a power transformer neutral as Position 3. These positions reflect some hierarchy in the surge protection of the substation equipment. From structural point of view the SA, which the SPS system is composed of, are power semiconductor components, designed as metal-oxide varistors intended to work under high voltages. In terms of reliability they are considered as non-repairable electronic devices. The SA reliability indices would take different values for the SA from different groups, depending both on the installation point and the degree of the voltage, power and temperature stress. Thus the failure rate estimations of SA in separate groups take different values. Typically the SA in a group are of the same type and have identical operational characteristics. For the purpose of this analysis and for to simplify the mathematical model chosen and easily to derive the most specific dependences, emphasizing at the same time the role of the SA characteristics and SA positioning, it is assumed that: All SA within groups installed at Position 1 are of a same type, are characterized by identical operational parameters and also by equal values of reliability indices; All SA within groups installed at Position 2 are of a same type, are characterized by identical operational parameters and also by equal values of reliability indices; The SA connected at the Position 3 are of a same type, are characterized by identical operational parameters and also by equal values of reliability indices; All groundings have equal values of its reliability indices at certain operational regime of the SPS; The voltage stress as well as the temperature stress of the SA of same type at same positions are equal at certain operational regime of the SPS. Applying the mathematical model describing the substation SPS reliability presented in [3] in the case when the number of OHL n = 3, the number of power transformers m = 2, and taking into consideration the assumptions above, it becomes possible this model to be simplified and it takes the form below
P t
17.G B . .t . 9. T1 . S1 6. T2 . S2 2. T3 . S3 B . e
(1)
where P(t) is the probability of SPS reliable operation for a certain period of time t, i.e. the SPS reliability function; λB denotes the basic failure rate of a metal-oxide SA, and its value is the same for all metal-oxide SA; λG denotes the failure rate of a grounding connection; T1 , T2 and T3 are coefficients, reflecting the temperature stress of the SA at Positions 1, 2 and 3 respectively; S1 , S2 and S3 are coefficients reflecting the voltage stress of the SA at Positions 1, 2 and 3 respectively; πφ is a
complex coefficient, which is equal to πφ = πC .πQ .πE [2], where πC is a coefficient reflecting the design peculiarity of SA; πQ is a coefficient reflecting the production quality level satisfied by SA; πE is a coefficient reflecting the environmental conditions during SA operation; t denotes the period of time for which the SPS reliability function is estimated. Let the substation equipment surge protection is performed by three SA types, e.g. Type 1, Type 2 and Type 3, different in their rated and operational characteristics, produced by same producer. Each of these SA types can be installed at the three possible positions identified above (see Fig.1). The task laid is the SPS reliability to be assessed for different variants of positioning the SA of the three types, and based on benchmarking of all possible variants an approach to selection of variants ensuring the highest level of substation equipment protection to be suggested. The number of possible variants is determined by the formula for the number of variations with repetition as below k
V p kp
(2)
where p denotes the number of positions; k denotes the number of SA types. In this case p = 3 and k = 3. Hence the total number of possible variants of the SPS 3
composition is V 3 27 . Let the basic characteristics of the three types of SA are as follows: The rated operational voltage UC is this, which a SA can endure long enough time without any damages, keeping its operational characteristics within the standard. Let the root mean square (rms) values of this, valid for the SA in the example are: UC = 84 кV for SA Type 1, UC=77 kV for SA Type 2, and UC=58 kV for SA Type 3; The protection level at which SA perform overvoltage limitation UP for the SA in the example its values are: UP=240 kV for SA Type 1, UP=226 kV for SA Type 2, and UP=212 kV for SA Type 3. For the SPS reliability assessment is chosen the methodology presented in MILHDBK-217F [2] as it is the most applied reliability prediction methodology for systems composed by electronic components [7]. According to this as well as to the Part stress analysis approach also presented in [2], the SPS reliability function for a certain time P(t), which is the basic reliability index of the SPS system, is determined by the reliability indices of the system components, i.e. the SA. The operational environmental conditions are specified in [2] as 14 types, depending on the location of the operation, climatic conditions, factors of the operating conditions severity, etc. There are presented there also initial reliability data regarding electronic components, necessary to perform a reliability estimation of system composed by such. Two different operational regimes are typical for an SPS. The first of it is the normal operational one. This is the case when the SPS and its components are put under normal stress in regard to operational conditions such as voltage, temperature, power dissipated etc. The SPS and its components can endure such operational conditions for a long enough time, i.e. for its lifetime.
The second operational regime is this when the SPS is put under extreme stress. This is occurred during a lightning stroke on an OHL followed by occurrence of an overvoltage wave. The duration of this regime is very short and usually lasts microseconds. In order to find out an applicable approach to proper SA set selection it is necessary the SPS reliability to be assessed during these two operational regimes. The reliability estimation is performed considering the respective reliability data regarding power electronic components presented in [2]. The indices and coefficient, valid for both of the regimes, take the values as follows: λB is equal to 0,0013.10-6 [h-1] as for transient suppresors/varistors; πC takes the value 1,0 corresponding to metallurgicaly bounded SA contact construction; πQ takes the value of 0,7 which corresponds to the highest quality level; πE takes the value of 8,0 , which corresponds to the outdoor operational condition. 2.1
SPS reliability function estimation during a normal operation
For SPS reliability estimation during a normal operation (Regime 1) it is appropriate to choose the SPS reliability function for a certain period of time as SPS reliability index for to estimate. Its estimation value can be used as a basis for benchmarking of different SPS compositions. The initial reliability data corresponding to this SPS operational regime are as follows: λG takes the value of 0,001.10-6 [h-1]; t is assessed as equal to 4320 h, which corresponds to the time between two consecutive periodic inspections and preventive maintenance, which typically is six months; The voltage stress is assessed as in [2], for operational linear voltage of 110 kV. The respective data regarding coefficients πT and πS are presented in the Table 1. Table 1. Values of πT and πS for the three SA types at the three positions, in Regime 1 SA type 1 2 3
2.2
T1 1,643920 1,696444 1,805500
S1 0,506767 0,626239 1,246741
T2 1,643920 1,696444 1,805500
S2 0,506767 0,626239 1,246741
T3 1,400415 1,446645 1,542765
S3 0,035108 0,043409 0,086372
SPS reliability function estimation during a lightning stroke
As this operational regime is a specific one, the SPS reliability analysis during this requires additional particularization. For this purpose a hypothesis is adopted. This is that an eventual lightning stroke hits an incoming OHL in close proximity to the substation. Practically this is the worst case of impact and this must provoke activation of the protective function of the entire SPS in regard to the basic facilities, such as the power transformers. This is called below Regime 2. As the duration of this SPS operational regime is extremely short, there are not any studies regarding uptime distribution of the SA at this. For to overcome this obstacle it is additionally assumed, that SA uptime distribution during this regime also can be described by the exponential random variables distribution and it can be applied the
same mathematical model as above. The difference here is that during the operational regime which is characterized by extreme stress is more sufficient to choose the probability of failure Q(t) of the SPS system as a basic reliability index for to estimate. At this SPS operational regime λG takes the value of 0,015.10-6[h-1]. The processes during a lightning stroke are analyzed by a computer simulation, presented in [4]. This is performed considering the conditions as follows: A lightning hits an OHL up to about 100 m far from the substation; The lightning stroke lasts 10 µs. This is the value of t; The substation is equipped with SA installed as it is already shown (see Fig.1). As a result of the analysis it is ascertained that the peak voltage value at the Position 1 is equal to 260 kV, at the Position 2 it is 201 kV and at the Position 3 its value is equal to 196 kV. Considering the respective voltage stress at each position, as well as the fact that the maximal junction temperature of SA in case of their activation is up to 800C [9,10], the coefficient values of πT and πS for this SPS operational regime are determined as it is shown on Table 2. Table 2. Values of πT and πS for the three SA types at the three positions, in Regime 2
T1
S1
T2
S2
T3
S3
4,438424 4,669512 5,033468
1,2147 1,4044 1,6420
4,216297 4,438424 4,788491
0,64991 0,75215 0,87830
3,231927 3,412731 3,698647
0,611381 0,706946 0,826331
SA type 1 2 3
The probability of failure of SPS is expressed as Q(t) = 1 - P(t).
3
Results obtained
The estimations of SPS reliability function P in Regime 1 and also the estimations of SPS probability of failure Q in Regime 2 for all 27 variants (Var. No) of SPS composition, are presented in Table 3. The variant type (V. type) is shown in the form x,x,x, where the numbers denote the SA type at the first, second and third position respectively. The dependence of SPS reliability function estimation P in Regime 1 and also the dependence of SPS probability of failure estimation Q in Regime 2 on the variants of SPS composition are presented graphically on Fig.2 and Fig.3 respectively. Table 3. Estimations of P and Q of SPS for 27 variants of compositions Var. No
V. type
Regime 1 P
Regime 2 Q.10-14
Var. No
V. type
Regime 1 P
Regime 2 Q.10-14
1 2 3 4 5 6 7
1,1,1 1,1,2 1.1.3 1,2,1 1,2,2 1,2,3 1,3,1
0,99953058 0,99952972 0,99952529 0,99948733 0,99948647 0,99948204 0,99926318
0,199840 0,222045 0,222045 0,222045 0,222045 0,222045 0,222045
15 16 17 18 19 20 21
2,2,3 2,3,1 2,3,2 2,3,3 3,1,1 3,1,2 3,1,3
0,99941718 0,99919833 0,99919747 0,99919305 0,99912951 0,99912865 0,99912423
0,244249 0,244249 0,244249 0,244249 0,266454 0,266454 0,266454
SPS relatability estimation P
8 9 10 11 12 13 14
1,3,2 1,3,3 2,1,1 2,1,2 2,1,3 2,2,1 2,2,2
0,99926232 0,99925790 0,99946571 0,99946485 0,99946042 0,99942246 0,99942161
0,222045 0,222045 0,222045 0,244249 0,244249 0,244249 0,244249
22 23 24 25 26 27
3,2,1 3,2,2 3,2,3 3,3,1 3,3,2 3,3,3
0,99908628 0,99908543 0,99908100 0,99886222 0,99886137 0,99885694
0,266454 0,266454 0,266454 0,288658 0,288658 0,288658
0,9996 0,9995 0,9994 0,9993 0,9992 0,9991 0,9990 0,9989 0,9988 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 26 27 Variants of SPS composition
Fig. 2. Dependence of P of SPS in Regime 1 on the variants of composition.
Probability of failure estimation Q.10-14
0,29 0,27 0,25 0,23 0,21 0,19 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 26 27 Variants of SPS composition
Fig. 3. Dependence of Q of the SPS in Regime 2 on the variants of composition.
4
Conclusions and final remarks
The data resulting from this study, presented above, cannot be used as highly accurate SPS reliability estimations. Their value is that they allow to be performed a benchmarking of all possible variants of SPS composition by SA with different characteristics. The results display that there are variants of SPS composition which provide the system with higher reliability than others. These are variants 1 to 6 from Fig.2 and 1 to 10 from Fig.3. These variants are characterized by roughly equal estimations of the SPS reliability function, and of the probability of failure respectively. Selecting a proper set of metal-oxide SA the maintenance engineers can apply the approach presented, in regard to a substation with certain n and m, considering SPS reliability and its estimations, valid for all variants of SPS composition, taking also
into consideration the technical and operational requirements regarding the basic substation equipment. During this process it is advisable the next steps to be followed: 1. Preparation of a list of all generally suitable metal-oxide SA on stock and/or available by request. Review of their referent characteristics; 2. Determination of the coefficients values and reliability indices values regarding each type of SA at each possible position, at each of the two operational regimes of SPS, considering SA referent and operational characteristics; 3. Reliability estimations of all variants of the SPS composition resulting from all possible variations in positioning of the SA of different types already listed. Benchmarking of all variants, based on estimations obtained. The current step is executed twice, i.e. once per each of the SPS regimes analyzed above. Selection of the SPS composition variants, providing the system with highest reliability; 4. Final choice of the SPS composition of metal-oxide SA on the basis of additional considerations, such as financial, operational etc. It is also advisable to be taken into consideration the maintenance experience of the personnel.
References 1. Holtzhausen, J.P., Vosloo, W.L.: High Voltage Engineering Practice and Theory, ISBN 978-0-620-3767-7 (2014) 2. Military Handbook, Reliability Prediction of Electronic Equipment, MIL-HDBK-217F, US Department of Defence, USA (1995) 3. Nikolov, N., Dimitrova, N., Georgiev, A., Vasileva, M.: Reliability Assessment of Electricity Distribution Substation Surge Protection System, Jubilee 40th International Spring Seminar on Electronics Technology, Sofia, Bulgaria, pp. 134-135, ISBN 978-619-706615-9 (2017) 4. Velikova, N., Dimitrova, R., Vasileva, M., Yordanova, M.: Model Research of the Protection against Arriving Atmospheric Surges in the Substation 110kV, International Scientific conference in Electro Energy, p.78., ISBN 978-954-20-0762-3 (2016) 5. Daman-Khorshid, H., Ajri, F., Shariatinasab, R.: Probabilistic evaluation of failure risk of transmission line surge arresters caused by lightning flash, IET Generation Transmission & Distribution, , 2014 - 8(2), pp.193-202 (2014) 6. Rausand, M.: System reliability theory, John Wiley & Sons Inc., New Jersey, ISBN 0-47147133-X (2004) 7. Nikolov, N., Papanchev, T., Georgiev, A.: Reliability Assessment of Electronic Units Included in Complex Electronic Systems, Jubilee 40th International Spring Seminar on Electronics Technology, Sofia, Bulgaria, pp. 132-133, ISBN 978-619-7066-15-9 (2017) 8. Bayadi, A., Harid, N., Zehar, K., Belkhiat, S.: Simulation of metal oxide surge arrester dynamic behavior under fast transients, International Conference on Power Systems Transients, IPST 2003, New Orleans, USA (2003) 9. Christodoulou, C.A., Assimakopoulou, F.A., Gonos, I.A., Stathopulos, I.A.: Simulation of metal oxide surge arresters behavior, IEEE Power Electronics Specialists Conference PESC 2008, Rhodes, Greece, pp. 1862-1866 (2008) 10. Miyakawa, Y. et al.: Influence of Temperature Variation On Characteristics of ZnO Elements, International Symposium on Electrical Insulating Materials - ISEIM 2008, pp. 119122 (2008)
