The medium voltage system performance analysis of front lightning discharges is very dependent on its modeling. As the model approaches the reality, the more it becomes extremely complex and time expensive. As a result, it generally leads to the adoption of some sort of simplifications and approximations.

The present work aims at the study of a large variety of effects of the lightning discharges, its impacts and preponderant factors for analysis in different real systems, as far as searching for a balance between the model's approximation and the resultant errors. With this in mind, it uses models that are more precise, stochastic process simulations, electromagnetic transient simulations, real information from the networks and statistical analyses. Therefore, it is possible to establish the main intervention points for the improvement of the medium voltage overhead distribution system performance front lightning
discharges.

Excerpt

The surge effects in the distribution networks are very

dependent on the lightning discharge current intensity, speed

of the return stroke and the point of impact, among others. The

two first parameters are relative to the lightning discharge

itself and, therefore, random; otherwise, the impact point is a

major influential factor for the calculation of the surges.

Table I show the peak induced surge voltages for lightning

discharges in a standard medium voltage distribution network

for the three phases. Each case represents a distance from the

network to the impact point of the lightning, where the first

case is closer and the last case is distant. It shows the little

influence of a three-phase over a single-phase analysis and the

influence of the distance in the induced surges.

TABLE

XAMPLE OF

IMULATION OF

VERVOLTAGE

S IN A

HREE

-P

HASE

IRCUIT

Induced Overvoltage's [kV]

Case

Phase A Phase B Phase C

49.60

48.80

49.60

24.70

24.50

24.70

15.00

14.60

15.00

9.12

9.00

9.12

5.58

5.52

5.58

2.88

1.74

1.56

1.74

0.78

0.72

0.78

0.48

0.18

Then, to reach a better analyses speed and minor modeling

complexity, a single-phase modeling of the circuits is used,

which is a valid approach that results in a faster analysis and

less time expensive process.

As well as to reach a balance between the model complexity

and the resulting error, it was adopted the use of a real

topology of the distribution network, using geographically

referenced database. In addition, to minimize the total error of

the simulations, many cases were simulated with the Monte

Carlo Method. However, some simplifications had to be taken

in order to reduce the complexity and the amount of

information required of the real topology.

For this reason, a plain area was adopted for the entire area

of the distribution network, that is, without any information

about topography or elevations. Also real elevated structures,

like towers, buildings and trees had not been considered

automatically; even so, they are simulated through the manual

inclusion of high points.

The adoption of these simplifications was mainly taken by

the difficulty and complexity of obtaining and dealing with

digital data about topography and elevated structures, since it

is necessary to implement the automated routines without

considering such simplifications.

Taking these simplifications led to a larger number of direct

interceptions from the lightning discharges for the distribution

networks, in other words, an overestimated result for the

worse case. However, through the manual inclusion of

elevated structures, some different cases can be simulated,

since a low natural shield for the network up to one high

degree of shielding, as a result of a larger number of induced

surges.

III. P

ERFORMANCE

IMULATIONS

Each performance simulation is executed in a great area,

covering the entire distribution network, however, not much

bigger than that, making possible to simulate a set of circuits.

It can deal not only with a typical urban network, more

uniform and with a greater network density, but also with rural

networks, more dispersed and non-uniforms, or even a mix of

both.

In this total area, and based on the number of discharges to

the ground by square kilometer per year (discharges/km²year),

adopted by the regional characteristic or an average, the

Monte Carlo Method is used to simulate 100 years or more of

lightning discharges. These simulations generate an enormous

mass of data that can be analyzed statistically, as a result

giving data that are more reliable for the studies.

The probable position of the discharge impact is defined

randomly within the total area, in the same way that the

current intensity, which follows a lognormal distribution with

an average of 31 kA, is random as also as the correlated speed

of the return stroke.

After these definitions, of the descending impact point and

current intensity, the verification of the attractiveness area is

initiated, following the Electro-Geometrical Model [1,10],

defining if that discharge will reach some structure, the

network or the ground.

With the impact point, current intensity and front time

defined for the lightning discharge, it is possible to calculate

the distribution network peak overvoltage. Leading to two

cases: When the discharge intercepts the network directly, the

peak overvoltage is calculated based on the parameters of the

traveling waves and surge impedance of the line. On the other

hand, when the lightning does not intercept the network, but

generates an induced surge, it is necessary to calculate the

distance from the impact point to the closer network point and

the electromagnetic fields, in order to verify the total

induction, as a result the surge peak overvoltage.

The method adopted for the calculation of induced surges

was the LIOV-EFEI [13, 14], this is based on the LIOV Code

[3] (Lightning Induced Over-Voltage) developed at the

University of Bologna and adapted, with some simplifications,

by the High Voltage Laboratory (LAT-EFEI) of the Federal

University of Itajubá, Brazil.

For the performance simulation, three major cases were

simulated: one rural network, one flat urban network and an

urban network considering elevated structures. The simulation

results for the flat urban network and for the rural are very

close, with minimal differences.

Because of the bigger density of branches and lines by a

small area, the urban network simulations lead to the rise of

the number of interceptions by lightning discharges, not only

by direct interceptions but also by induced surges. However,

the statistical results in terms of current intensities and

overvoltage's are practically the same.

However, when the elevated structures are simulated, in

different sets, it is evident the decrease of the amount of

lightning intercepting directly the network. In fact, the

presence of closer high points to the networks exerts a greater

or minor degree of shielding of these to the direct

interceptions. In the case of metropolitan regions, which have

a bigger amount and concentration of high constructions,

towers, buildings as many other elevated points, the level of

direct interceptions of the network is extremely low.

A. Rural Simulation Results

The rural network, Fig. 1, by possessing a much greater area

than the urban, needed a large amount of lightning discharges

simulations. A total of 707,214 discharges were simulated,

equivalent to 364 years for the case of an average of

6 discharges/km²year and 323.85 km² of total area. From this

total, only 4.34% had directly intercepted the network, this is

caused by the sparse distribution of the circuit in a big area,

leading to a small probability of interception by the network of

a lightning discharge.

Fig. 1. Adopted rural medium voltage distribution network real topology.

For the 30,693 discharges of direct interception, the

histogram of the current intensities is shown in the Fig. 2.

From the fitted lognormal distribution, it was obtained the

current intensity value with 50% of probability: 40.24 kA;

with 90% probability: 16.66 kA and with 10% probability:

97.22 kA.

Fig. 2. Current intensity histogram for the lightning discharge direct

interception into the medium voltage rural distribution network.

Fig. 3. Induced overvoltage's histogram for the medium voltage rural

distribution network.

In the case of the discharges that had not intercepted the

network directly, that is, the remaining 676,521 occurrences,

they had generated the overvoltage histogram of the induced

surge, as it can be seen in the Fig. 3. In which it is possible to

see the low values found, that were caused by the general

great distances from the impact point to the network. From the

fitted lognormal distribution, it was obtained the value of

surge with a probability of 50% of occurrence: 6.46 kV; with a

probability of 90%: 1.28 kV and with 10%: 32.61 kV.

B. Urban Simulation Results

In the case of the urban network without the presence of

elevated structures, Fig. 4, it was simulated 286,237 lightning

discharges, what is equivalent to 2,141 years of lightning

discharges in the network. A number achieved by the total

area of 22.28 km² and the regional average of

6 discharges/km²year.

Fig. 4. Adopted urban medium voltage distribution network real topology,

without the presence of elevated structures.

From this total, 95,392 discharges intercepted the network

directly, where, by the fitted lognormal distribution, it was

obtained the current intensity with 50% of occurrence

probability: 38.14 kA; 90% of probability: 16.81 kA, and

10%: 86.53 kA, as it can be seen in the Fig. 5.

Fig. 5. Current intensity histogram for the lightning discharge direct

interception into the medium voltage urban distribution network, without

elevated structures.

The remaining 190,845 discharges that did not intercept the

network represent more than twice of the number of direct

interception. The Fig. 6 represents the overvoltage's

histogram, where the distribution identification could not fit

well, caused by the histogram non-uniformity.

Fig. 6. Induced overvoltage's histogram for the medium voltage urban

distribution network, without elevated structures.

As a result, the distribution that got optimum performance

in the test was the tri-parametric Weibull, in which it is

possible to obtain the overvoltage value with 50% probability

of occurrence that was 29.49 kV; the 90% probability:

6.46 kV and the 10% probability: 77.62 kV.

The maximum and minimum values had been similar to the

found ones in the previous rural case, however, as it can be

seen, the great majority of the cases of the surge for the rural

network stayed below 30 kV.

C. Urban with Elevated Structures Simulation Results

To make the simulated urban circuit better represents a real

case of an urban area, it is necessary to take into consideration

the presence of elevated structures, as trees, buildings, high

constructions, towers, among others. Thus, through the

inclusion of simulated elevated structures in the circuit, which

attracts the lightning discharges, diverting them from the

networks, it will supply a shielding effect.

Tests had been made with two configurations for the

simulations of elevated structures. Firstly, the structures are

enclosed with an average distance of 10 m from the network

and with 40 m between two consecutive structures. In this

configuration, a high index of network shielding was

achieved, proved by the index of 2.24% direct interception by

the network [5,7].

Secondly, the structures had been inserted with an average

distance of 5 m for the network and 80 m between two

consecutive structures, reaching a lower degree of network

shielding, with an index of 15.43% of direct interception of

the network [5,7].

These two previous simulations had shown a behavior of a

dense and a sparse urban region, simulating an urban zone of a

central region of a great city and a small town, respectively.

To simulate an average case, that could contemplate both the

situations, it was adopted the inclusion of elevated structures

with an average distance of 10 meters of the network and

70 meters between two consecutive structures, Fig. 7,

searching for an intermediate condition between the

previously simulated.

Fig. 7. Adopted urban medium voltage distribution network real topology,

with the presence of simulated elevated structures.

In this condition, 259,250 lightning discharges had been

simulated, or the equivalent to 1,939 years of lightning

discharges reaching this region. From these, only 2,938 had

been intercepted directly by the network. On the other hand,

elevated

structure

the

ground

intercepted

256,312 discharges, which is equivalent to 98.87% of the

cases. The shielding of the network was relatively high,

however, the study continued for the good statistical

representation of this case and by being a better study

condition for the induced overvoltages.

In the case of the direct interceptions, the Fig. 8 shows the

histogram of the found current intensities. The value of

discharge current intensity with a probability of 50% of

occurrence was obtained by the fitted lognormal distribution

and was 11.60 kA; for the case of probability occurrence of

90%: 5.29 kA, and for 10% probability: 25.44 kA. It is easy to

see the fall in the direct interception current intensities

observed, from 38.14 kA to 11.60 kA in the occurrence

probability of 50%. Due to the fact of the bigger intensity

discharges, by possessing a greater attraction distance, that

reaches firstly an elevated structure than the network

conductors.

Fig. 8. Current intensity histogram for the lightning discharge direct

interception into the medium voltage urban distribution network, with

simulated elevated structures.

The Fig. 9 shown the histogram of the induced

overvoltage's, in this figure is also difficult to identify the

distribution that can better fit, as a result it was used the one

that has the better adherence in the tests, which was the tri-

parametric Weibull distribution. Where can be obtained the

values of surge overvoltages with a probability of 50% of

occurrence: 31.45 kV; of 90% probability: 13.70 kV and of

10% probability: 72.20 kV.

Fig. 9. Induced overvoltage's histogram for the medium voltage urban

distribution network, with simulated elevated structures.

The increase in the found surge levels is notable, where the

maximum limit increased from 160 kV, in the case without the

presence of elevated structures, to 375 kV in this in case. It

was due to the high-intensity discharges that reach elevated

structures closer to the network. However, as in the previous

case, the biggest concentration was in the band of the 10 to

30 kV, what resulted in the values of probability very similar

between the two cases. Where for the 50% probability it was

increased from 29.49 to 31.45 kV, and 90% probability from

6.46 to 13.70 kV, however, the 10% probability had indeed a

small decrease from 77.62 to 72.20 kV.

IV. T

RANSIENT

IMULATIONS

Electromagnetic transitory simulations through the ATP

program had been executed to detail and analyze the dynamics

and distribution of the surge into the system and the response

from the installed equipment.

For this reason, 45 cases of induced overvoltage's due to

lightning discharge had been simulated, plus 45 more

simulations for overvoltage's generated by direct impact

lightning discharge, totalizing 90 simulations. All of them

executed in an urban feeder, in which 36 of them, that is, 40%

had been executed in the presence of elevated structures. This

was done because of the fact that the results of an urban

network without the elevated structures possess similarity with

the results of a rural network, consequently we have that 54 of

the cases can be attributed to a rural or a low-density urban

network and the remaining to a dense urban region.

The simulations take into care the line surge impedance,

surge arresters, medium voltage transformers and its basic

insulation level, insulators and the line critical flashover

overvoltage for the entire distribution network under analyses.

The surge source was modeled with the 2-slope ramp

model, the lines with the lumped series R-L-C parameters

obtained from the Line Constants Routine. The transformers

with the high-frequency capacitive- model obtained from the

real transformers measurement with the Schering Bridge. The

flashover by voltage controlled switches and the surge

arresters by the pseudo-nonlinear resistance modeled by the

voltage by the current standard curve for gapped silicon

carbide (SiC) surge arresters, as shown in Fig. 10.

This was made necessary for the, up till now, great amount

of units installed in the field of surge arresters with the older

technology gapped silicon carbide, being then possible to

compare the results.

Fig. 10. Voltage by the current non-linear curve for a standard gapped silicon

carbide (SiC) surge arrester for medium voltage distribution system class

15 kV.

The Fig. 11 shows the histogram of the simulation results

for the current dissipated by the surge arresters under the

network direct interception lightning discharge. In this

histogram, it is possible to see that the majority of the surge

arresters dissipation stayed between 1.5 to 10 kA.

On the other hand, Fig. 12 shows the histogram of the

simulation results for the terminal overvoltage in the

transformers generated by network direct interception

lightning discharge. From this histogram, it is possible to

notice that the medium voltage transformer class 15 kV, are

receiving surges bigger than its basic insulation level, between

a 100 to 500 kV.

Fig. 11. Surge arresters current dissipation histogram for direct interception

of the lightning discharge.

Fig. 12. Distribution transformers, 15 kV class, medium voltage terminal

overvoltage's histogram for direct interception of the lightning discharge.

In the opposite case, when the surge arresters and

transformers are under induced surges, it is notable the lower

intensities. As it is possible to see in Fig. 13, all of the current

dissipated by the surge arresters are under one kA.

Also for the transformers overvoltage's, we can see in

Fig. 14 that all of them are under the minimum transformer

basic insulation level of 95 kV for the 15 kV class.

In other words, when the distribution systems are in a dense

urban region, with many elevated structures that can deviate

and dissipate the lightning discharge, there is no problem or

fault occasioned by the induced surges in the transformers or

even in the surge arresters. In fact, not even the insulators are

requested, because none of the cases provoked surges bigger

than the lines critical flashover overvoltage. In these cases, the

worse problem will be the transferred surge to the low voltage

circuit, which cannot deal with that amount of energy.

Fig. 13. Surge arresters current dissipation histogram for induced surge.

Fig. 14. Distribution transformers, 15 kV class, medium voltage terminal

overvoltage's histogram for induced surge.

However, if the distribution system was intercepted directly,

a common situation in rural or less dense urban regions, the

surges achieve high values that are capable of causing a

transformer, an insulator, or even a surge arrester to fail. Table

II shows the probability of a direct interception lightning

discharge to cause a failure in a transformer, according to its

basic insulation level for the 15 kV class, or to cause an

insulator flashover, according to the lines critical flashover

overvoltage.

TABLE

ESULT

UMMARY FOR THE

IMULATIONS OF

IRECT

IGHTNING

ISCHARGE

NTERCEPTION

Condition

Probability

> BIL 95 kV

80.0%

> BIL 110 kV

75.6%

> CFO 150 kV

46.7%

> CFO 175 kV

35.6%

> CFO 200 kV

26.7%

> CFO 225 kV

22.2%

Table III summarizes the occurrence probability of

transformers overvoltage and surge arresters current, obtained

from the fitted distribution from all the simulations.

Although, as it considers half of the cases as a direct

interception and the other half as induced surges, it must be

properly pondered with the bigger probability of the induced

surges, which could indeed decrease the values shown. Since

there are 45 cases of direct interception, it should have

300 cases of induction, or 7 times more, considering a direct

interception probability of 15%.

TABLE

III

ROBABILITY

UMMARY FOR

LL OF THE

IMULATIONS

Probability

Transformer

Overvoltage

[kV]

Surge

Arrester

Current [A]

341.2

7,394.0

236.8

5,683.2

65.3

1,710.8

18.0

260.1

12.5

126.6

95.0

------

110.0

------

By doing a simply pondered adjustment of Table III, the

"corrected" values of Table IV are achieved.

TABLE

ROBABILITY

UMMARY

DJUSTED

ALUES

Probability

Transformer

Overvoltage

[kV]

Surge

Arrester

Current [A]

116.21

4,612.70

90.79

2,514.12

38.00

39.94

15.91

0.001

12.43

95.0

------

110.0

------

V. F

IELD

ALIDATION

Through a research and development project in partnership

with AES Sul Brazilian utility, approximately 300 gapped

silicon carbide surge arrester unities had been removed from

the same distribution urban network of the simulations. These

units were submitted through an analysis technique for the

measurement of the bigger current intensity that the surge

arrester had discharged; this is done based on the electrodes

etchings measurement and comparison with laboratory made

marks. From this great data volume, the graph of Fig. 15 was

traced, in which it is possible to verify the great proximity and

similarity between the simulated and real measurement cases.

It is clearly in Fig. 15 that, mainly for the current intensities

with 50% probability or greater, both curves present a great

similarity, in fact, only below the 40% probability that the

curves diverge a little. From the graph, we can notice that the

occurrence of current intensities of 700 A or lesser possesses

probability equal or superior to 50%. For a 10% probability,

we have that the current intensity situates in the band of 5 to

6 kA, for a current intensity of 10 kA the probability will be

approximately 5%, and for a current intensity of 20 kA the

probability will be approximately 2%.

Fig. 15. Field fitted distribution vs. simulation fitted distribution comparison.

VI. C

ONCLUSIONS

The effect studies of the lightning discharges on the

electrical systems are complex and time demanding, because

they involve a lot of factors and modeling. The proper process

of the discharge is very difficult of being explained, for this

reason, generating a wide gamma of modeling possibilities,

each one adapted to the necessity of a specific study.

In addition, the simulation process of the effect of the

lightning in the distribution systems is sufficiently complex by

the great amount of influence factors, such as the network

topology, installed equipment, system shielding, installed

protection, elevated structures in the neighborhoods, amongst

others.

To avoid all these difficulties, some rational attitudes are

necessary to reach a simplified level in the process without

changing the simulation results from the practical values,

keeping it with an adopted tolerance. This implies in the

research and development of models that can reach a balance

between the modeling complexity and the results similarity,

further statistical simulations, through the Monte Carlo

Method, are made to reach a good response from the

simulations.

Through the development of many studies in the High

Voltage Lab. of the Federal University of Itajubá (LAT-EFEI),

Brazil, the LIOV-EFEI code is developed for the calculation

of induced overvoltages in distribution networks [2,3,13]. This

is based on the LIOV code developed at the University of

Bologna, however, adapted to the conditions in which it would

be used in the LAT-EFEI, generating an extremely practical

and efficient code.

Moreover, the development of a simulation routine became

necessary, which the use of the stochastic process. This

routine has as objective the effect simulation of the lightning

discharges on the medium voltage overhead distribution

systems, by the direct interception of it or through

electromagnetic induction. Thus an entire simulation program

was developed, which uses geographic databases with all the

pertinent information of the network, beyond user data and the

Monte Carlo Method.

With the use of the real data from the network, statistical

procedures and some approaches, the simulation routine

reached a high capacity of generating cases with an excellent

similarity with field results. Through these system

performance simulations for lightning surges, it is possible to

figure the improvement possibility of the system protection.

Where, for predominantly rural systems, the surges by

lightning direct impact stayed below 4% probability, with an

average current intensity of the order of 40 kA and

overvoltage of the order of 6 kV.

On the other hand, for the less dense urban network, that is,

small towns, where the number of elevated structures is small,

it has a direct impact index with 15% of probability. However,

for an urban network of the metropolitan region, with a higher

density of constructions, an index of only 2% of direct impact

was achieved. The result for an average case had presented

around 10% probability of direct interception by the

distribution system, with the average current intensity of the

order of 12 kA and average overvoltage intensity in order of

31 kV.

Consequently, it is thus evidenced the importance of the

system type to be analyzed, regarding the lightning discharges.

Therefore, systems in rural regions and different systems in

urban regions present very different results. However, it is

obvious the importance of better studies of the surges

provoked by electromagnetic coupling over the ones caused

by direct interception, where the average probability stayed

around 90% of the cases being by induction.

Analyzing the results of the electromagnetic transient

simulations it is possible to observe that, for induced surges:

none surge arrester dissipated current superior than 1 kA; none

transformer was submitted to an overvoltage superior to its

basic insulation level (BIL) and the occurrence possibility of a

flashover in the insulators only could be verified in 2 cases,

that is, 4%.

As this represented around 90% of the cases, it is evident

that the medium voltage overhead distribution systems would

not be subjected to failures or interruptions caused by induced

surges. However, to this be an absolute truth, the operational

condition of the surge arresters, insulators and the

transformers withstanding must be in satisfactory conditions.

Indeed, through laboratory tests on transformer unities, even

new ones, it was possible to notice that it is not true for the

majority of the cases.

It is also important to remind the necessity of the presence

of surge arresters in each medium voltage terminal of every

transformer installed in the system and even through the

network. A condition that, not rare, is verified inadequate or

for the lack of units or by miss-specification. As a result, we

have that the no observance of all the above requirements,

mainly for the transformers and surge arresters, let the system

into situations of failures, faults or interruption even in the

cases of the lightning-induced surge.

Although for the simulated cases of direct interception, the

situation becomes critical. Since it has that: the average

current dissipated by the surge arresters goes up for an average

of 9 kA; the transformers overvoltage's surpassed the

withstanding in 80% of the cases and the flashover probability

in the insulators was next to 50%.

Table V contains a summary of the electromagnetic

transient performance simulations of the system, in the cases

of surges by induction and direct impact.

TABLE

UMMARY OF THE

ESULTS OF THE

RANSITORY

IMULATIONS OF

LECTROMAGNETIC

Surge Arresters

Discharge Current [A]

Transformers

Overvoltage [kV]

Surge

Min.

Avg.

Max.

Min.

Avg.

Max.

Induced

283

892

Direct

Impact

680

4,779 48,412

214

2,173

It is possible to notice by the values presented in the Tables

II, IV, and V that for the condition of surge caused by direct

interception, the probability of failures and interruptions in the

system could be very high. In the surge arresters, the

dissipated current reached the mark of 48 kA, intensity

superior to the capacity of almost all of the surge arresters for

distribution networks class 15 kV. For the transformers, the

fault probability stayed between 70 and 80%, as the average

overvoltage value of 214 kV is very superior the biggest

standard BIL of 110 kV. As for the flashover probability in the

insulators, for the standard condition of 150 kV of critical

flashover overvoltage (CFO), it achieves 50%, however, this

value can be reduced to 20% with the increase of the lines

CFO to 225 kV. Therefore, for the direct interception surges in

the system, the possibility of an interruption is almost certain.

To validate the simulations, a comparison of the results with

the current intensities dissipated by the gapped silicon carbide

surge arresters units removed from the field was made,

Fig. 15. In which, the similarity of the simulations with the

real cases can be verified, occurring a little distortion for the

probabilities lower that 40%, where the simulations had

presented bigger values than the real cases. These differences

could be attributed to the approach used and the

simplifications adopted, which tend to present a small error

regarding the physical event that they model, consequently,

the sum of these errors cause the verified discrepancy, indeed

leading to an overestimated result, the worse situation.

Finally, some best practices for the utilities are shown

below:

A. Surge Arresters

It is very important to attempt to the correct specification

and quality of the surge arresters, as well as searching for the

best cost versus benefit in its use (quantity and displacement

through the network). Where, through studies of insulation

coordination, it is possible to achieve the best distribution of

the units, reaching a balance between the maximum protection

and the minimum of units.

It was demonstrated that, for rural systems, the surge

arresters must have the capacity of withstanding the maximum

current of 40 kA. On the contrary, for urban systems, it is

more than appropriate to work with units that have a

maximum capacity of only 10 kA. Since the network is

entirely composed of overhead lines and the transformers has

an assured quality and withstanding capacity.

B. Insulators

It is evident that raising the insulators BIL, and

consequently the line CFO, is possible to reduce significantly

the possibility of flashovers and subsequent failures in the

occurrence of surges.

C. Transformers

To get the best protection condition of the transformers,

preventing failures by lightning discharges, it is essential that

they always were specified with the biggest BIL of its class.

Beyond always using surge arrester unities, correctly

specified, in each of its terminals.

It is recommended that regular statistical verification of the

new unities should be made, to certify and assure its quality.

One of the tests that can better assure the quality of a

transformer is the voltage impulse test.

CKNOWLEDGMENT

The authors gratefully acknowledge the support given by

the Prof. Carlo Nucci of the University of Bologna, Italy,

helping in the development of the LIOV-EFEI code. The

authors also gratefully acknowledge the support technically,

financially, and the supplied information by the AES Sul

utility, Brazil, special thanks to the M.Sc. Eng. Hermes R. P.

M. de Oliveira.

EFERENCES

[1] Marco A. M. Saran, "Lightning Overvoltage's in Medium Voltage

Lines", Master Thesis, Federal University of Itajubá, Brazil, Feb. 2009.

[2] Manuel L. B. Martinez, Pedro H. M. dos Santos, "Study of the Induced

Voltages in Distribution Networks, Guide for the Performance

Improvement of the Overhead Distribution under Lightning

Discharges", High Voltage Lab., Federal University of Itajubá, Brazil,

March 2004;

[3] Carlo A. Nucci, Mario Paolone, "Calculation of Induced Voltages in

Medium Voltage Overhead Systems due to Lightning Strokes Using the

LIOV Code", Report for the Second Phase of the R&D Project for the

AES Sul Utility, October 2003;

[4] Marco A. M. Saran, Rafael R. Bonon, Manuel L. B. Martinez, Hermes

R. P. M. De Oliveira, Carlo A. Nucci, Mario Paolone, "Performance of

Medium Voltage Overhead Distribution Lines Against Lightning-

Induced Voltages: A Comparative Analysis", GROUND'06 e 2nd LPE -

International Conference on Grounding and Earthing & 2nd

International Conference on Lightning Physics and Effects, Maceió,

Brazil, November, 2006;

[5] Marco A. M. Saran, Manuel L. B. Martinez, Hermes R. P. M. De

Oliveira, "Performance of Medium Voltage Urban And Rural

Distribution Lines Front Lightning Discharges And Induced Surges",

GROUND'06 e 2nd LPE - International Conference on Grounding and

Earthing & 2nd International Conference on Lightning Physics and

Effects, Maceió, Brazil, November, 2006;

[6] Marco A. M. Saran, Rafael R. Bonon, Manuel L. B. Martinez, Hermes

R. P. M. De Oliveira, Carlo A. Nucci, Mario Paolone, "Performance of

Medium Voltage Overhead Distribution Lines Against Lightning

Discharges", International CIGRÉ Symposium TPLEPS Transient

Phenomena In Large Electric Power Systems, Zagreb, Croatia, April

2007;

[7] Marco A. M. Saran, Manuel L. B. Martinez, Hermes R. P. M. de

Oliveira, "Performance of Medium Voltage Urban and Rural

Distribution Lines Front Lightning Discharges and Induced Surges",

15th International Symposium on High Voltage Engineering, Ljubljana,

Slovenia, August 2007;

[8] Marco A. M. Saran, Manuel L. B. Martinez, Carlo A. Nucci, Mario

Paolone, Hermes R. P. M. de Oliveira, "Performance Analysis of

Medium Voltage Overhead Distribution Line Against Lightning", 19th

CIRED, International Conference on Electricity Distribution, Vienna,

Austria, May 2007;

[9] Marco A. M. Saran, Manuel L. B. Martinez, Carlo A. Nucci, Mario

Paolone, Hermes R. P. M. de Oliveira, "Comparative Performance of

Medium Voltage Overhead Distribution Lines Designs Submitted to

Induced Voltages", Power Tech, Lausanne, Switzerland, July 2007;

[10] IEEE Guide for Improving the Lightning Performance of Electric Power

Overhead Distribution Lines, IEEE Std 1410-2004, T&D Committee,

IEEE Power Engineering Society;

[11] John G. Anderson, Thomas A. Short, "Algorithms for Calculation of

Lightning Induced Voltages on Distribution Lines", IEEE Transactions

on Power Delivery, Volume 8, Number 3, Pages 1217-1225, July 1993;

[12] Parameters of Lightning Strokes: A Review, Lightning and Insulator

Subcommittee of T&D Committee, IEEE Transactions on Power

Delivery, Vol. 20, No. 1, January 2005;

[13] Pedro H. M. dos Santos, "Performance Analysis of Medium Voltage

Circuits Front Induced Lightning Impulses", Master Thesis, Federal

University of Itajubá, Brazil, March 2007;

[14] Ricardo G. de Oliveira Jr., "Induced Voltages in Medium Voltage

Lines", Master Thesis, Federal University of Itajubá, Brazil, August

2008;

[15] Andrew R. Hileman, "Insulation Coordination for Power Systems",

Marcel Dekker Inc., 1999;

[16] G. Vernon Cooray, "The Lightning Flash", IEE Power Series, Volume

34, 2003;

[17] Lou van der Sluis, "Transients in Power Systems", John Wiley & Sons,

2001;

[18] Mustafa Kizilcay, "Power System Transients and Their Computation",

Osnabrück University of Applied Sciences, Germany, 2000;

[19] Protection of MV and LV Networks against Lightning, Joint CIGRE-

CIRED Working Group C4.4.02, 2005;

Marco Aurélio M. Saran received his B.Sc. and M.Sc. degrees in electrical

engineering from the Federal University of Itajubá, Brazil.

Currently, he is a Consultant and Researcher of the High Voltage Lab. of

the same university, dealing with research and development projects, tests on

electrical equipment, among other activities. He is involved in the electric

power system area, especially insulation coordination, reliability and

distribution systems. He is author or co-author of several papers and technical

reports.

Manuel Luiz B. Martinez received his B.Sc. and M.Sc. degrees in electrical

engineering from the Federal University of Itajubá, Brazil and his Ph.D.

degree in electrical engineering from the São Paulo University, Brazil.

Currently, he is Full Professor in power systems of the Federal University

of Itajubá and the Main Professor of the High Voltage Lab. at the same

university. He is involved with the electric power system area and his research

interests include high voltage, electromagnetic compatibility, power systems

transients, reliability, insulation coordination, electrical distribution,

transmission and energy efficiency. He is author or co-author of several

papers, technical reports and journal publications.

Frequently asked questions

What is the main topic discussed in this text?

The text discusses the effects of lightning discharges on medium voltage overhead distribution systems, focusing on simulations and performance analyses of these systems under lightning strikes.

What parameters are considered for lightning discharge simulations?

The simulations consider parameters such as lightning discharge current intensity, speed of the return stroke, point of impact, ground flash density (discharges/km²/year), and the use of the Monte Carlo Method to simulate many years of lightning discharges.

What are some simplifications made in the modeling process?

Simplifications include adopting a plain area without topography or elevation data, not automatically considering elevated structures like towers and trees (though they can be included manually), and using a single-phase modeling approach for faster analysis.

What is the Electro-Geometrical Model used for?

The Electro-Geometrical Model is used to determine the attractiveness area and whether a lightning discharge will strike a structure, the network, or the ground.

What are the two main cases considered for calculating peak overvoltage in the network?

The two cases are: direct interception of the network by the lightning discharge, and induced surges caused by lightning strikes near the network.

What is LIOV-EFEI code?

LIOV-EFEI is a code based on the LIOV (Lightning Induced Over-Voltage) code developed at the University of Bologna, adapted by the High Voltage Laboratory (LAT-EFEI) of the Federal University of Itajubá, Brazil, to calculate induced surges in distribution networks.

Which types of networks are simulated in performance simulation?

The simulations include rural networks, flat urban networks, and urban networks considering elevated structures.

How does the presence of elevated structures affect the performance of urban networks?

The presence of elevated structures, such as buildings and towers, tends to decrease the amount of lightning directly intercepting the network by providing a shielding effect.

What is the typical probability of direct interception in metropolitan regions?

Metropolitan regions, with a higher amount of high constructions, tend to have extremely low levels of direct interceptions by the network.

What role do surge arresters play in the distribution network?

Surge arresters dissipate the surge energy to prevent overvoltage's in the electrical circuits of the system such as transformers and insulators.

What was the purpose of performing electromagnetic transient simulations through ATP?

Electromagnetic transitory simulations through the ATP program were executed to detail and analyze the dynamics and distribution of the surge into the system and the response from the installed equipment.

What are some of the best practices for utilities regarding lightning protection?

Best practices include correctly specifying and maintaining the quality of surge arresters, raising the basic insulation level (BIL) of insulators, and ensuring that transformers have the highest possible BIL, as well as always using surge arrester unities in its terminals.

What was validated to determine the accuracy of simulations?

Real measurements of current intensities released by surge arresters were compared with simulated outcomes to validate the reliability of the model.

According to simulations, which network type experiences the greatest chance of outages due to lightning?

Overhead distribution systems in rural regions are more susceptible to faults and outages brought on by induced surges, according to simulation data.

Excerpt out of 9 pages - scroll top

Details

Title: Analysis of Overhead Distribution Lines Performance under Lightning Surges
Course: Power Systems Electrical Engineering
Authors: Marco Saran (Author), Manuel Martinez (Author)
Publication Year: 2009
Pages: 9
Catalog Number: V338848
ISBN (eBook): 9783668287969
ISBN (Book): 9783668287976
Language: English
Tags: analysis overhead distribution lines performance lightning surges
Product Safety: GRIN Publishing GmbH

Quote paper: Marco Saran (Author), Manuel Martinez (Author), 2009, Analysis of Overhead Distribution Lines Performance under Lightning Surges, Munich, GRIN Verlag, https://www.grin.com/document/338848

Analysis of Overhead Distribution Lines Performance under Lightning Surges

Excerpt

Frequently asked questions

What is the main topic discussed in this text?

What parameters are considered for lightning discharge simulations?

What are some simplifications made in the modeling process?

What is the Electro-Geometrical Model used for?

What are the two main cases considered for calculating peak overvoltage in the network?

What is LIOV-EFEI code?

Which types of networks are simulated in performance simulation?

How does the presence of elevated structures affect the performance of urban networks?

What is the typical probability of direct interception in metropolitan regions?

What role do surge arresters play in the distribution network?

What was the purpose of performing electromagnetic transient simulations through ATP?

What are some of the best practices for utilities regarding lightning protection?

What was validated to determine the accuracy of simulations?

According to simulations, which network type experiences the greatest chance of outages due to lightning?

Details