YT YouTube2Text

YouTube2Text → Transcripts → Insulation coordination – Lightning transient overvoltages

Insulation coordination – Lightning transient overvoltages — Transcript

by EMTP · 8,518 words · 1,525 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:02hi and welcome to this
0:04video on installation coordination the
0:07lightning part
0:11so in this video we are going to
0:14see how to perform an insulation
0:16coordination
0:17in emtp and especially
0:20and focus of course on the lightning
0:25so before watching this video i highly
0:28recommend you to take a look
0:29at the other recorded video which is
0:31called
0:32origin and classification of
0:34overvoltages
0:36i will do some reminders
0:39here but to fully understand the content
0:42of this video it's
0:43highly recommended to watch this this
0:45first one
0:47okay so well first of all if you uh
0:51the first reminder actually from this
0:52video it's about
0:54the different type of overvoltages there
0:57are like
0:59very fast transients over voltages which
1:03are actually the fastest then you have
1:05the fast front over voltage
1:07which are lightning over voltages you
1:09have switching
1:10over voltages and temporary overvoltages
1:14which are
1:15the slowest like typically a fundamental
1:18frequency over voltage like a single
1:20phase ground fault or
1:22load rejection for resonance
1:25why it is important again to repeat what
1:28is important to
1:30classify those over voltages it's
1:32because
1:33an apparatus will
1:36waste them a certain level according to
1:40the shape of your voltage
1:43for lightning or voltage the
1:46the surge is very steep and so you will
1:49have
1:49here the time from zero to peak which is
1:52very short
1:53in this condition the apparatus will
1:56withstand
1:57and higher over voltage compared to
2:01a switching over voltage where the time
2:03to peak is much
2:04slower and so it means that the maximum
2:07level
2:08which for lightning is called a bil
2:11basic impulse level the
2:15this level is higher than for a
2:18switching level
2:21okay so here you can see approximately
2:23the dynamic of this
2:24search for lightning the classification
2:28so basically when we are going to do a
2:30an analysis
2:32we do a simulation or even if you have
2:35something
2:36a real lightning strike on field
2:39we kind of assume that the shape of the
2:42overvoltage
2:43we will have will be close enough
2:47to the lab test all equipment
2:50are doing all the tests
2:53and equipment have passed through and so
2:56that the level
2:57uh the bil which was obtained by lab
3:00measurements
3:01to determine the bil we assume that
3:04for the any shape of lightning strike we
3:07will get
3:08during the simulation or on field will
3:10be similar and so the same level
3:12can be assumed
3:17okay so uh there are several type of
3:21lightning over voltages uh first of all
3:23you have the induced overvoltage
3:25and this is mostly uh this mostly
3:28affects
3:29the distribution network it's when the
3:30lightning strike does not strike
3:32directly the line but something close by
3:35and you have an overvoltage on the line
3:39for that there are some tools actually
3:40we have one here which is called earlier
3:43liov which is used in this purpose but
3:47that's not what we're going to do today
3:50then you have what is called
3:53the lightning strike so when the
3:56lightning strike
3:57on a shield wire so you have the tower
4:00and at the top you have two ground wire
4:01which are called a chill wire
4:03those are designed to capture lightning
4:07and so you may have over voltages when
4:09the voltage rises
4:10so high on those chill wire that you
4:13have
4:14a flash over between the ground and the
4:17face
4:17and because it is because between the
4:19ground and the face it's called a back
4:21flashover
4:22and so that's typically what we study
4:24when we do empty simulation mtp
4:27simulations or
4:28we study this case the backlash we also
4:31will apply
4:31a lightning strike on the tower top
4:34on the chill wire
4:38in some region of the world where the
4:42the lightning density isn't so high
4:45um it's not economic it doesn't make
4:47sense economically to
4:49put those shielding wires or to
4:52perfectly protect a line very often you
4:54will need two chilling wires
4:56so often you will put only one or none
4:58according to
5:00again some economical uh consideration
5:04and if you don't have so many lightnings
5:06on the region and so in this case it may
5:08happen that the lightning strikes
5:10straight on a phase without striking
5:14the shielding this is cause this is
5:16called a direct strike or
5:18shielding failure so this is also
5:20something you have to study in emtp and
5:22make sure
5:23the overvoltage
5:26there is no too bad of a voltage for
5:28that
5:30typically we don't study lightning
5:33tracking inside the substation
5:35because we assume that the substation is
5:37protected and you have lightning mass
5:38which will capture the lightning
5:40and so the electrical operators are
5:43safe inside the substation so when you
5:46do the emtp simulation
5:47you have a device which is a lightning
5:50strike which we will see later in the
5:52video which is a current source
5:53you will connect this current source on
5:56the incoming transmission lines but not
5:58inside the substation
6:02so the objective of any mtp study here
6:06there will be several things if you talk
6:08about transmission line itself
6:10you will come out with what is called
6:11the line performance
6:13which is the flash over rate of an
6:16overhead line
6:17okay you have a transmission line how
6:19often
6:20you will have a back flashover or
6:23shielding failure
6:26and so this is typically like a number
6:28ones per
6:29100 year 200 years and so on
6:32it can be much less or much more so this
6:35is for transmission lines
6:37and then if you talk about equipment
6:40inside substations now we are going to
6:42determine the rate of failure
6:46if you have so the first step will be
6:47you will have a back flash over on the
6:49line
6:49and then this back flashover will create
6:51an overvoltage on
6:53phases which will propagate to the
6:55substation and may damage
6:57equipment okay so there we will see in
7:00this
7:01video the procedure to determine that
7:04and finally it's also to determine uh
7:06the survivor ratings
7:08um so the switching the lightning
7:14surge
7:16the impulse level of the the third
7:19arrestor but also the position of the
7:21sort arrester
7:22and this is crucial for lightning um
7:26one thing you have to remember which is
7:28important is that for lightning
7:30the substation is actually a circuit and
7:32you don't assume that the voltage is the
7:34same at each point in the substation
7:36which means that if you have a
7:38substation connected to a transformer
7:39but there is a conductor in between
7:41during a lightning strike the voltage at
7:45the
7:45at the point of connection of the third
7:47arrestor will not be the same as the one
7:49of the point of connection of the
7:51transformer because the lightning strike
7:52travels through the
7:53station propagates through the
7:56substation
7:58all right so some definitions many of
8:00them i've already discussed
8:02uh so you have the first thing which is
8:04called a direct lightning strike
8:06so it's when the lightning strike
8:08directly on the face
8:10of the transmission line without hitting
8:12the shielding wire
8:15the back flashover it's a flashover
8:17between
8:18ground and phase so when the lightning
8:20strike
8:21the neutral wire the top of the tower
8:24and you have
8:25i mean at the top of the tower or in
8:27between it doesn't matter
8:29um you will have the flash so you the
8:31voltage will rise
8:32then the voltage difference between the
8:35ground and the phase wire will be high
8:37and you will have a flashlight
8:38between so this is a backlash then you
8:41have the back flashover rate
8:43which is the number of back flashover of
8:45the line
8:46per 100 kilometers so you take a section
8:49of line
8:50per year so this will depend on the on
8:52the tower geometry
8:54and on the ground density of lightning
8:57in the region the grounds of like
8:59the ground density is how often you get
9:02lightning strikes
9:03for example if you are in uh you know in
9:06the
9:07louisiana region in the u.s here you
9:09have a very high lightning
9:11lightning strike ground density whereas
9:14if you are
9:15more up north you know in the west coast
9:17of canada it's much
9:20lower
9:22okay then we will see together how to
9:24determine what is called the critical
9:26current
9:27the critical current it's when you will
9:29apply a lightning strike during the
9:30simulation
9:31on at a certain point on the incoming
9:34line so typically you will have the
9:35substation
9:36all the incoming line the incoming line
9:39and you will put the lighting strike
9:41closed uh on the line close by the
9:44substation
9:45within few miles or even less you know
9:47hundreds of
9:48meters or miles you will put the
9:51lightning strikes here
9:53and the critical current will be the
9:55minimum
9:57peak current of this lightning strike
9:59which will
10:00cause a failure in the substation
10:04so for that you will have to know of
10:06course the wisdom label
10:08of all equipment which are called the
10:11bil
10:12the basic lightning impulse insulation
10:16level
10:18so you will need to know those values
10:20simulate the lightning strike monitor
10:22the voltage
10:23at at each point in the substation and
10:27see if the voltage reach
10:28this maximum level and if it's the case
10:31then
10:31you are above the critical current and
10:34so it will be
10:35we will run several simulations in order
10:37to determine the critical current at
10:39each point
10:40of the of this incoming
10:45actually line some point when you go
10:46further and further away from the
10:48substation
10:48this critical current become very high
10:50it can be more than 200 kilo amps
10:54at some point when it's over 200
10:56kilometers we
10:58are going to standard we assume that
11:00it's not possible anymore
11:02and so that's that means that such a
11:04lightning strike does not exist it's too
11:06high
11:06or it's uh the probability is much too
11:09low so you
11:10you you will say that at this point no
11:12lightning would
11:14uh no no lightning would create the
11:17damage in the substation
11:18and this this points this uh
11:21distance from the substation to
11:24on the line where you have you reach a
11:27critical current of 200 kilo amps
11:29it's called the critical distance we
11:32will see that uh
11:33earlier okay so we discussed here the
11:37bil the bsl
11:39um so the bil the definition is
11:42um it give a probability of
11:46the wisdom probability of 90 percent
11:50however you will see for if you watch
11:53the video of the switching
11:54insulation coordination it's a little
11:56different for lightning
11:58because basically the the level of
12:00lightning for
12:01the difference between ninety percent of
12:04winston and fifty percent of wisdom it's
12:05almost nothing
12:07and so typically for lightning because
12:09the statistical distribution of the
12:11width and voltage is so narrow
12:14even for internal or external insulation
12:17we will use
12:21just one a single level which will be
12:24the level to design or design
12:26and if you get an overvoltage above this
12:27level then
12:29you get you get a damage and all
12:32flash of backlash so it will be a bit
12:36easier than
12:37compared to switching where you have
12:39some equipment like insulators which are
12:41called non-self-restoring insulation
12:45and which have a which has a statistical
12:47behavior and forced us to
12:49cross different statistical roles
12:54okay so the limit distance is actually
12:56what i call before the critical distance
12:58i apologize the real name is limit
12:59distance
13:00so again this is when you apply
13:03lightning strikes on the line
13:05you find the critical distance and at
13:07some point
13:08the the critical current will be higher
13:10than 200 kilo amps so this is
13:13this point here is the limit distance we
13:14will repeat that again
13:17later to make sure it's clear
13:22okay so then you have what is called the
13:23maximum shielding
13:25current uh we will we will discuss
13:29what is the max maximum chilling current
13:31in the next slides
13:34just a quick introduction to that when
13:36you will see you have
13:37your phase conductor where they get
13:39protected by a shielding current
13:42and you will see the theory but
13:45the bottom line is for low
13:50lightning currents you are more likely
13:52to get a chilling failure than for
13:54a high lightning current and actually
13:57when the lightning current is very high
13:59you most of them you don't have shading
14:02failure
14:03and so the limit between that is the
14:04maximum shielding current it's the
14:06highest magnitude of lightning current
14:08which may cause a shielding failure
14:11so we will see the theory right after so
14:13don't worry if it's not so clear
14:24all right then you have the
14:25representative lightning stroke current
14:27which is the
14:28minimum value of lightning current at
14:30specific point of impact which produces
14:33over voltages that the equipment has to
14:35withstand
14:39right so then like i said you will do
14:42this study and you will
14:43come out with some line performance
14:46or so for line performance the outcome
14:49of the study will be the number of flash
14:51over
14:52per 100 kilometers and per year then for
14:55substation it will be
14:57the mean time between failure and so you
14:59have here some
15:00typical performance criterion here for
15:04ehv and hd system or here for
15:07different type of substation of course
15:10of course here there is an economical
15:12problem
15:13uh like it's a trailer between
15:15economical concern
15:17and failures um in order to increase
15:20the mean time between failure and the
15:22substation either
15:23you get equipment with higher bil
15:27which costs of course more which are
15:29more expensive
15:30or you put more surge arresters
15:33or you change the tower design so all
15:36that doesn't cost
15:37when you have one failure every 500
15:40years
15:41you know that's why you made me wonder
15:43why would i spend more money on that so
15:44it's really like a trade-off between
15:47economical concern and
15:52design
15:55okay so again how we are going to
15:59do that in the mtp so the first step
16:00will be to model completely the design
16:03so let's take a simple example where we
16:05have one substation with one incoming
16:07line
16:08so we are going to
16:12place a lightning strike
16:15for current source at the top of this
16:17line and at several points going further
16:19and further away
16:20for each point where we apply this
16:23current source
16:24we will we will vary
16:27the um the current the
16:30lightning strike peak current so we'll
16:33vary it in order to find
16:35the limit between failure or no failure
16:38if i have a peak current lower than this
16:42point i have no failure if i have a
16:43bigger
16:44higher than this point i have a failure
16:46so that's again the critical current
16:48you will be supposed to do that at each
16:50point on the line but of course to
16:52simplify it what we're going to do is to
16:53cut the line into section
16:56um very often what is done is we do that
16:58at each tower top
17:00but it's also common to do it at the
17:02tower top and at the middle of the line
17:04and so that's how you will
17:05sectionalize your line and apply the
17:08lightning strikes
17:09just because you cannot apply
17:10continuously it will be much
17:12way too much cases and it's not so
17:16relevant
17:18okay so after the mtp study you will
17:20have the critical
17:21current at each tower
17:25so
17:29if you if you do actually lightning on
17:31the cable system
17:32so of course the lightning will not
17:33strike the cable system but
17:35it may for example strike on the
17:37overhead line and then propagate
17:39through the cable system through um
17:42or through transformers or in this case
17:45it's the voltage between sheets and
17:47ground which
17:48which may cause the the flashover you
17:51may want to monitor also the arrestor
17:53energy
17:53so that's all the outputs of any of the
17:57mtp simulation
18:02okay so the lightning the lighting we
18:05have different lighting model in mtp
18:08the one recommended by standard is the
18:10sea gray model
18:12so the secret model has a certain shape
18:14um
18:15some people do studies with double
18:18exponential
18:19um it's not necessarily recommended to
18:21use that because
18:23um double exponentials are actually the
18:26the the surge the current surge which i
18:29use for testing
18:31uh but uh according to c grade iec
18:34it's recommended to use this remodel for
18:37lightning strike but
18:38both can be used the two models are they
18:40are here in the entp
18:45okay so now um
18:51okay so now let's discuss a little bit
18:53the tower geometry
18:55so remember i in the introduction i
18:58explained that
19:00when you have a lightning strike
19:01according to the lightning intensity
19:04you may hit the chilling wire or the
19:07phase and when you have a lower um
19:11low intensity you're more likely to hit
19:14the the face and have a shielding
19:17failure
19:18so why is that well it's because
19:21when you have a conductor or
19:24any metal metallic structure which is a
19:28conductor
19:29you have an attractive radius the
19:32attractive radius is
19:33a radius if a lightning strike enter it
19:37the lightning strike will end up on the
19:39metallic structure
19:41so having this uh chilling wire or
19:44sometimes you have two i think those
19:46create kind of like a cylinder
19:48over the line and if the lightning
19:50strike until the cylinder
19:52it goes into the shielding wire and this
19:56circle this uh attractive radius
19:59depends on the the lightning current
20:03intensity
20:04and so for low intensity
20:08for low intensity lightning strike you
20:10will have a smaller radius
20:12whereas for high intensity lightning
20:14strike you have a very large radius
20:16and that's why this large radius of the
20:18on the conductors at the top of the
20:20tower we basically cover the entire line
20:22and protect
20:23completely the line i will show you a
20:24drawing after which will
20:26and repeat the explanation
20:41okay so then the the so the lightning
20:44distance
20:45is the distance between an object on
20:47earth and a lightning
20:48leader within which the lightning
20:51must approach before the
20:55point of strike is determined
20:59and so for this um for for this
21:03attractive radius you have several
21:04models that exist in the literature
21:07um some of them depend on the current
21:10some of them depend also on the height
21:12so it may depend on both height and
21:15current
21:19okay so those are important factors to
21:21consider when the tower will be designed
21:23and
21:23determine again the the
21:26the mean time between failure underlying
21:28performance
21:32so also if the goal is to determine
21:36how much lightning is going to strike
21:38your line okay so
21:40try to imagine to see a single the line
21:43from the top you have the
21:44chili wires and those ceiling wires
21:47create
21:48a cylinder an attractive radius so a big
21:51cylinder
21:52and when you know the regional brand
21:53density so how much lightning strikes
21:56per year
21:57and you see this the ground cover
22:00from this line and seeing from the top
22:02that's how you can calculate how many
22:04lightning will strike your line per year
22:06and of course this will depend again on
22:08the lighting density because the
22:09lightning density will unlock
22:11the larger radius
22:16okay so let's put a drawing behind all
22:17that so it's easier to understand
22:20here you have your transmission line
22:22with phase a
22:23b and c here you have two chilling wires
22:28then you have here your attracted ray
22:31attractive radius
22:33okay so you see the attractive radius
22:36here you have
22:37the equation which is yours actually
22:39which will be used here for the
22:41for the webinar today which is the ritz
22:43from april
22:44and this one depends on the height and
22:47the current
22:49okay so assume any lightning strike for
22:52example 100
22:53amps lightning strike so you have i you
22:56have the tau the height of this
22:57conductor and that's how you calculate
22:59the radius
23:00then you have this radius here you do
23:02the same thing called the second shield
23:04wire
23:05and you have this second radius here do
23:07the same thing for the phase
23:09and you have here the radius around the
23:10face
23:12for a low current here you see that you
23:14will have a possibility
23:15the lightning lightning leader is coming
23:17from the top here of course coming from
23:19the top
23:20if it enters this zone which is the
23:22striking distance
23:23it will end up in the on the conductor
23:26whereas if you enter this zone it will
23:29end up there
23:30and if the lighting and density is very
23:32high
23:33rc is going to increase and so here b
23:36will end up coming in top of a
23:39and so then you will have no chance for
23:41any lightning
23:42to go to uh to go on the on the face and
23:45so you will have no chilling failure
23:49okay so this is called the um
23:53whatever all right so yes so that's um
23:58on the geometry point of view so that's
23:59what is in poland so in a
24:01in a country where i don't have so much
24:03ground uh
24:04while you have a very low height where
24:07you have a very low
24:09um ground density of lightning
24:12you will have only one chili wire at the
24:14in the middle
24:15and so basically if you have a very high
24:17lighting strike you will have very
24:19large radius which will protect your
24:20line but for lower
24:23level of lightning strike sometimes you
24:26will have chilling failure but it's okay
24:27because a
24:28lower level of lightning intensity and
24:31so they may not cause a damage
24:33that's the whole point of doing the
24:34study here
24:44okay so if if we repeat a little bit the
24:47procedure here
24:49and for that i will pop a
24:52an emtp example
24:56okay so here you have a simplified model
24:58of substation and we come back to that
25:01we never apply a lightning strike on the
25:03substation because we assume the
25:04substation is protected by
25:07lightning mass tents which will capture
25:09every lightning possible
25:11trying to strike the substation so what
25:13here
25:14that is the end of the substation on
25:15this side and so we this tower is still
25:18protected in the substation
25:19so we start applying lightning strike on
25:22the line
25:24here in our case we will try to apply on
25:26each tower so here you see three
25:28lightning strikes but it's actually one
25:29after the other
25:30so let first so we can forget those two
25:33here and actually delete them so first
25:37we
25:38we apply on the first hour we vary the
25:41intensity
25:42so we have no failure in here so we
25:45monitor the voltage
25:47uh if the overvoltage during the
25:48simulation remain below the bils it
25:50means there is no failure if it goes
25:52above there is a failure it means there
25:53is a failure
25:54so we'll vary for example for this point
25:57we will get
25:5875 kilo amps for failures
26:01but so if this is the limit if i apply a
26:04lightning strike below 75
26:06there is no failure and if i apply a
26:09lightning strike above
26:1175 there is a failure now it means 75 is
26:15my critical current
26:17then i go and i will assume it's the
26:19same critical current for all
26:20this section for the line here and for
26:22half the line here
26:24then i will place this lightning strike
26:27on the next
26:28next hour and do the same thing because
26:31i'm further away
26:32here now the like the um the critical
26:35current will be a bit higher
26:36maybe 100 and i would assume it's the
26:39same critical current for all that
26:41if you don't like this assumption you're
26:43free to apply the lightning in between
26:45and cut this line into section and apply
26:48the lighting in between
26:49and i will go like that step by step
26:51until the critical current is above 200
26:54kilo amps
26:54then like we said we assume this is a
26:58case which has
26:58the probability which is too low to be
27:01considered and so
27:02now we have our critical distance so for
27:05example let's assume
27:06here the critical current is 210
27:10okay then it means we don't consider
27:12this tower and
27:14the critical distance will be the length
27:16of all those sections those spans
27:18up to this half line
27:22okay so that will be the first step to
27:23find the critical current at each point
27:26then what do you do with that
27:30then you determine the number of
27:31children failure
27:33so it will depend of several things the
27:35ground
27:36density the what we call here
27:41in ng the regional ground flash density
27:45multiplied by l1 l1
27:48is the length of the line so the total
27:52length of the line
27:53and um gc is
27:56the exposure distance here actually i
28:00sorry i wasn't so cute and i got decided
28:03myself as well
28:04here what in this slide what we show is
28:06how to determine the shielding failure
28:09so the changing failure the goal here is
28:12to find
28:13the maximum current which may create
28:16achieving failure
28:18okay so you understand that as you
28:19higher i
28:22the the radius here increases and at
28:24some point d
28:25becomes zero so that's the goal here is
28:27to find
28:29what is the maximum current which may
28:31create a
28:32shielding failure so that's how we do
28:34that
28:35we find um we calculate basically here
28:41the the limits so the limit will be i m
28:45i am for which there is no d
28:48equals zero and then we'll do an
28:50integral of
28:52the length of the line multiply by d
28:55which is here multiply by two of course
28:58because it's
29:00two sides and um
29:03and then the uh the so the lens
29:06d multiplied by two and the ground
29:08density so you have this equation here
29:10you do the integral from
29:13the critical current to
29:17the critical current to the maximum
29:19chilling failure and that's how you get
29:21the shielding failure rate
29:24okay so the critical current is
29:26calculated
29:27um it's calculated by emtp so you get
29:31those critical current the way we
29:32explained
29:34and you can do that for the line the
29:36critical current here
29:37will depend on the section so you may do
29:40this per section and get the chilling
29:42failure
29:42per section and add them up together
29:48as a if you if you'd like to
29:51simplify your life you may also here
29:54calculate
29:55the chilling current like that if you
29:57know the third impedance of the line
29:59you may calculate here you may just use
30:03this calculation and
30:04without doing the simulation in empty
30:08here when we determine the shielding
30:10failure it's actually for the line
30:12performance
30:13so that's why in this calculation we
30:15don't reconsider the substation
30:17we consider u50 which is the critical
30:21flash over the critical flash over
30:23voltage of the line
30:25and so we just assume the failure on the
30:27line okay so when you
30:29have this chilling failure rate you only
30:31focus on the line
30:33then the goal is to focus on the
30:34substation you
30:36consider the critical current
30:40of which will create the flashover in
30:42the substation
30:46okay so here by design if you want
30:48perfect shielding then you have to set
30:51here the critical current equal to i m
30:53or
30:54you have to as um you have to arrange
30:57the geometry of the line so i
30:58am which is the value of intensity
31:02where d is zero to make sure that
31:06i am equal i c
31:17all right so now let's quickly discuss
31:19how to
31:20model everything in emtp so first of all
31:24a lightning study is something very
31:27local
31:28because it is local
31:31you don't need to model the entire
31:34system of the country to study lightning
31:37it's mostly you will
31:41you will model everything in the radius
31:44of a few miles or few kilometers
31:46close to the substation where you want
31:48to analyze lightning
31:50so if we take a look quickly here at an
31:52empty model
31:54you have the substation here we assume
31:55there is only one incoming line to
31:57have a simplified case but in reality
31:59you may have zero
32:00and you see here we have several steps
32:04and at some point we don't model anymore
32:08when we stop modeling there is one thing
32:11important to do
32:12when you apply your lightning on the
32:14tower for example let's say here you
32:16move this lightning and you put it
32:20at this point when you apply the
32:24lightning you know the lightning strike
32:25will create
32:26a search a search voltage the search
32:29voltage will travel through the line and
32:31reflect
32:32of course here if you stop modeling here
32:35you have to make sure when you play the
32:36lightning strike that it doesn't come
32:38here and
32:39come back that it doesn't go here and
32:40come back so what we typically do for
32:43that
32:44is we will here put a section of
32:47transmission line which is
32:49long enough so the lightning never comes
32:51back
32:52so typically you will put 50 kilometers
32:54or 20 kilometers and that should be
32:56enough
32:57the second strategy is to use a negative
32:59voltage source with
33:00an impedance which is equal to the surge
33:04impedance of the line
33:06but again to simplify in my opinion the
33:08easiest is to take the same transmission
33:10line here but
33:11generate it for a very long uh
33:14a very long section of 20 or 50
33:17kilometers
33:18and when so when you do the simulation
33:20the lightning strikes goes there and
33:22never have time to come back before the
33:23end of the simulation
33:26okay so that's how you will model
33:27boundaries in your system
33:29uh you put an ideal voltage source
33:33it's a good idea also to adjust here the
33:36phase
33:36angle to get worse case right so
33:39basically
33:40when you do the simulation uh you do one
33:43simulation and you
33:44you adjust your voltage source to get
33:46the maximum
33:47overvoltage in the substation and then
33:49you keep them like this
33:53so this is something you do at the
33:54beginning then once you have determined
33:56what's the worst case
33:58you just keep it like that
34:07all right then the for transmission
34:09lines um
34:10it's recommended to use a frequency
34:12dependent model or even better white
34:14damp models
34:15so i to uh to model them i don't go too
34:18much into detail we have specific videos
34:20on how to create transmission lines
34:23one thing here is that you see we model
34:25each span
34:26individually so you have
34:29so typically for a switching study or
34:33other type of study with modeled
34:34entire line in one device
34:38but here it's really each section so you
34:40will have
34:41the three phases and the chilling wires
34:44in this example we have only one shield
34:45wire
34:46you see it's here number one and then
34:48you have phase a b
34:49and c and so we repeat this section
34:53several times
34:54so that's how you do that then you also
34:56need to model the tower
34:58we'll come back on the tower model later
35:00but we will have to do it
35:02basically for that we use a lossless
35:05propagation line
35:07with a footing resistance we'll see
35:08after which is a cp model
35:12very often actually when you try to
35:14create a frequency dependent model
35:16or a wideband model for a very short
35:18section of line so one span which is you
35:20know 100 meters or so
35:22or sometimes more sometimes often it's
35:25very tough
35:26for the frequency dependent algorithm to
35:28fit that
35:29and you it may not be able to fit it or
35:32may require
35:33a very very high number of poles to do
35:35that and create numerical problems
35:38if that happens one thing that can be
35:39done is instead of
35:41generating a frequency dependent model
35:43of line you enter all the power
35:45configurations so the position of each
35:47conductor
35:48on space and all that but instead of the
35:51running the fd you generate
35:52a constant parameter line and you
35:56generate that for uh 50 kilohertz
36:00so the parameters resistance and
36:01backhands capacitance are
36:03evaluated for 50 kilohertz
36:07it's it's okay in one sense because i
36:09mean it's trying to work like that
36:11because the lightning strike is
36:12very high frequency the other frequency
36:15presents during the study is the syst
36:17the 60 hertz or 50 hertz frequency which
36:20for such a speed it looks like a dc
36:23voltage in here so that's why it's fine
36:25to do that
36:28if you study if the if you want to be
36:31very precise
36:32you may also model the corona effect
36:36the corona effect will give a capacitive
36:40behavior to your line which will
36:43lower the stiffness of the
36:46of the overvoltage so it's a good idea
36:48to use that if you want to be precise
36:50especially for very high voltage
36:52if you don't you will be conservative so
36:54it's fine as well
36:58okay so for air gap it's uh it's
37:01important to model the air gap
37:03of course because that's what will
37:04create the
37:06the flashover so when the air gap is
37:09short
37:10you can assume that as soon as there is
37:12a
37:13flashover it's like an ideal switch
37:16closing
37:17whereas if it's a long air gap more than
37:19one meter then it's
37:21important to reproduce the leader
37:22propagation which we add more delay and
37:25add an impedance to this
37:26arc
37:30okay now how do we model the substation
37:33itself
37:34everything must be modeled basically
37:36each section of conductor
37:38must be modeled if it's more than 10
37:40meters
37:41so approximately 30 feet then you should
37:44use a cp line
37:46so for the cp line we will enter the
37:48characteristic impedance
37:50and the length of this line then you can
37:53assume
37:54a traveling wave
37:57the propagation speed
38:01of about 80 of the speed of light
38:06this is for air insulated substation for
38:08gas insulated substation it may be
38:10slightly different
38:14then if you have a conductor shorter
38:16than that
38:17then simply use a pendant
38:20with a typical value of one micro henry
38:23per meter
38:26for gis typically the
38:30surge impedance is about 60 to 80 ohm
38:34for those sections of line whereas for
38:36air insulated substation it could be 150
38:38200
38:42then you have to place the surge
38:44capacitance of
38:45every single equipment if you have a bus
38:48bar
38:48you have to put the street capacity of a
38:50transformer a
38:52ct open breaker or clothes breaker
38:56every single equipment has to be
38:57represented by a three capacitance
39:00you can find typical value of three
39:01capacitance in is standard
39:04we actually have a device which is
39:06called a stray capacitance which is a
39:08library
39:09database or stray capacitance so you can
39:11use this device as well
39:16okay so it's the same for transformers
39:19to
39:19it's possible actually to be even more
39:21precise for transformer and to use
39:23a high frequency model of transformer
39:26so if you have done a frequency test
39:28where the
39:29the open circuit impedance of the
39:32transformer is measured for a wide range
39:34of frequencies
39:35then you may get this impedance data
39:37impedance that source frequency and
39:39create
39:40a frequency dependence impedance to
39:42reproduce
39:43your transformer so that's uh that's
39:46something
39:46feasible often it's not done for
39:50for lightning studies because it's a lot
39:51of work but to be precise it's possible
39:53to do that
39:57for such arresters it's very important
39:59here to use a frequency dependent model
40:01of soldier resistor
40:02so there is a specific video on how to
40:05create those frequency dependent model i
40:07will come back to them
40:08quickly after but yeah keep that in mind
40:11and the lightning source also is a sig
40:14resource
40:15the lightning current source is a secret
40:16source
40:20okay so at the end of the day if you see
40:21a substation like that it will look like
40:23this where you have
40:25any time you have here a conductor you
40:27will model it with
40:28a cp line here even very short
40:31conductors were modeled with cp9 three
40:34meters it's okay
40:35it will force to use a very small time
40:37step
40:38to simulate that but it's extra
40:40precision
40:42you see that each equipment are modeled
40:44by a capacitance
40:46and you have surgery so that's
40:53okay so the lightning strike here you
40:55can find it
40:56if you go to um
41:00if you go to sources here you can grab
41:03what is called seed currency so this is
41:06our lightning strikes
41:08even better what i would recommend you
41:09to do is to use this example which is
41:12available on our exchange platform you
41:15can
41:15contact the technical support to get it
41:18and
41:19to use this model because in this model
41:20you only enter
41:22the uh the peak current of lightning
41:25strike for example here 10 000
41:28and automatically it will adjust all
41:31other parameters so the stiffness and
41:34the tail duration
41:36so it's it's pretty nice you don't need
41:37to
41:39you don't need to calculate or latch
41:40yourself
41:46okay so it's uh the current source looks
41:47like this so you have several parameters
41:50the peak current uh you have the
41:53time the tf here which is uh
41:56well you can see what it is in the in
41:58this figure
41:59uh you have uh the stiffness
42:02this is kind of important and a major
42:04difference compared if you do the
42:06lightning strike using this model
42:08compared to a double exponential model
42:10uh the stiffness will not be the same
42:12and as you know the stiffness of current
42:14has a huge impact on overvoltage so
42:17that's why
42:18it's recommended to use this model
42:23also typically in the lightning strike
42:25model we put a resistance in parallel
42:28and that's what we've done here in this
42:29little model you have
42:32the secretable with this in parallel
42:44okay so here you have some equations on
42:47how to calculate
42:49uh current stiffness and all the
42:51parameters
42:53um if you hear the actually that's the
42:56the summary of it so the current
42:58the stiffness that tells you the
43:00calibration and
43:01the time to get some t
43:05t30 i don't remember what it is
43:12it's a tf
43:15okay so you have everything you need
43:17here i'm not sure about td30 but
43:19what you need really is i sm th
43:22and i would have to double check if t30
43:27is
43:27tf not sure about that it's actually
43:30written in the
43:32in the script of this device here
43:35let's see if we can quickly find out
43:38find that out till team
43:58not sure if uh td30 is actually used in
44:01the model oh yeah it is td30
44:11no i don't think td sort is actually
44:13used in the model itself
44:15it's just another parameter
44:18apologize i don't remember what it is
44:20exactly
44:23all right so then um then you need to
44:26model the tower
44:28okay let me keep going on the tower so
44:30the tower you will approximate that it's
44:32a cylinder
44:33and then you will model you will
44:36calculate
44:37the surge impedance based on
44:41the parameter of this of this cylinder
44:44okay so the average radius uh the
44:46section
44:47here h1 h2 if the
44:50radius is if the radius is changing
44:59okay so then what the way you are going
45:01to model that is to use a cp line
45:04in the cp line you put the surge
45:06impedance the propagation speed of 80
45:1080 of speed of light and that's how you
45:12will model this tower
45:16let me show you how it is
45:21okay so you see here that's our tower um
45:25before actually showing you that let me
45:26show you a little example a drawing
45:28you see you have a tower here and so you
45:30have different
45:32sections for each sections here
45:36and the arms are also model however
45:38because the arm are very short
45:41we only use inductances for that it's up
45:44to you
45:44you may use cp line or inductances you
45:46may do the calculation the same
45:48calculation for the arms
45:49it's fine to use conductances else
45:52typically it's a one micro henry per
45:54meter
45:55for those sections okay so this
45:58is our transmission line here you have
46:00all the faces so
46:01phase a b and c for example and the top
46:04here is the tower top so the chilling
46:06wire
46:06that's where we'll apply the lightning
46:08strike so when we apply the lightning
46:10strike you see that it will travel
46:11through the line
46:13it may create an overvoltage between
46:14here the a phase and an overvoltage and
46:17this is the air gap model where
46:19you may have a flashover and here you
46:22have the
46:23grounding pin of the of the tower where
46:26we will
46:27model the grounding resistance
46:30okay so the grounding resistance is
46:32modeled as such
46:34when you have a lightning strike you
46:36have a lot of current going
46:37through the ground inside the ground and
46:39these create what is called ionization
46:42which will vary the resistance
46:45the grounding resistance and this is
46:48here the equation
46:49provided by a triple e for the grounding
46:51resistance
46:52so you have here it depends on the
46:55lightning density and
46:56ig and ig the parameter calculated which
46:59depends
47:00on the solar resistivity mostly and
47:05the footing resistance of at 60 hertz so
47:08this may be determined
47:10by measurements and the solution
47:13gradient so here we typically take this
47:15value 400.
47:17how do we model that now well we just
47:19create a small control circuit
47:21we use this controlled uh
47:24norton equivalent where we set i
47:28the current at zero the whole time so
47:30basically it's just a control resistance
47:33and you see we measure here the current
47:36then we have the equation i just show
47:39here
47:40you the input u2 which is the footing
47:43resistance multiplied by square root of
47:46two
47:47one plus the current u1 divided by
47:5118 500 so that's exactly
47:55this equation here
47:59okay and so this will control the
48:02resistance
48:03and here we have a small selector when
48:06the intensity
48:07is lower than 18 500
48:10then we consider instance is the
48:14footing resistance and when it's higher
48:16we
48:17use this uh this calculation
48:20okay so it's just a selector okay so
48:23that's how we model that
48:24again this tower model is available
48:27on demand so if you have a lightning
48:29study to do you simply use that and
48:31modify it
48:35this example is actually inside the mtp
48:37example
48:44right so now let's discuss how do we
48:47model those insulators
48:50like i said you have two possibilities
48:53which will mainly depend on the length
48:55of the
48:57insulator there is one the first method
49:00is
49:01when you have a voltage time curve which
49:03means that the insulator will be able to
49:05withstand a certain voltage for a
49:07certain time only and after this time
49:09you will have the backlash over
49:10so the breakdown you may have this curve
49:13here
49:14then one way you can do that is again
49:17the same way we've done for the
49:18resistance
49:19uh use a control um
49:23use a control circuit to determine that
49:26so you will enter those points
49:28in a table and measure the current
49:32create a timer and when you reach the
49:35curve
49:35basically close the gap it's possible to
49:38do that but it's
49:39not really what we recommend it's
49:41possible to do it but
49:43i will show you after other way to do it
49:45which are easier
49:47so the other model which is recommended
49:49actually
49:50is for gaps which are lower than
49:541 meters or 1.2 meters
49:58so here it uses an equation to determine
50:00if there is a flashover or not
50:02so it's used area criterion model so
50:05according to the length
50:07the the size of your gap okay you have
50:10those parameters and you can interpolate
50:12in between
50:14and then automatically the model can
50:16continuously integrate
50:18the difference between v-gap and v0 so
50:21the voltage
50:22in the in the gap in the at the
50:25insulator terminal
50:27versus v0 here and when these integral
50:31becomes larger than d
50:32it's have a closed ideal switch so the
50:36current will now go through the
50:38uh goes through this gap and the voltage
50:41on the phase
50:42and the neutral and the ground will be
50:44the same
50:46okay so this model is valid for um
50:49it's valid for a smaller gap and you can
50:52find it here in the library
50:55switches and it's called
50:58um air gap
51:02all right so that's the model we use
51:04here in the slime model
51:05it's right there
51:12okay the second model to use you can use
51:15it's for
51:16larger gaps more than one meters more
51:19than one meter
51:20so this one is uh now reproduce the
51:23leader propagation
51:24okay so it will actually take when there
51:27is a flashover it will take some time
51:29for the flashover to go from this point
51:30to here and
51:31there will be a resistance associated
51:35with that
51:36so here the equation is written there
51:39the only input for this model will be
51:41the lens and the mall is already built
51:44in emtp
51:45so you can simply grab it from the
51:47library
51:48you have here an example if you go to
51:51application
51:52air gap leader okay the model is there
51:56and you can select here the lens
52:02okay so again this one is recommended to
52:03be used when you have a lens high
52:05higher than one meter
52:09okay so then you have the software store
52:11model
52:12here again we did a i repeat we did a
52:14complete video
52:16on how to model such arresters for
52:19lightning
52:20it's important to use the frequency
52:22dependence of resistor
52:24so the frequency dependence of the
52:25restore is model like this as a
52:27combination of
52:28rlc component and
52:32two nonlinear resistance which have an
52:34exponential
52:35fitting there is a procedure in order to
52:39build that
52:40we actually have a device which does it
52:42automatically for you
52:44so the only things you will enter are
52:46the uh
52:47the impulse test
52:51results so you have what is called v10
52:54which is the voltage for a lightning
52:57serve of 10 kilo amps
52:59the lightning surge takes eight
53:01microseconds to go to the peak
53:03and 20 microseconds to go to the top so
53:06it's a
53:07820 so you need this value
53:11and you need also a switching search so
53:13a switching surge which
53:15has you know for example a 30 60 so
53:18third takes 30 seconds to go from zero
53:21to the peak
53:22and then 60 microseconds sorry
53:24microseconds for the tail
53:26so you need those two uh informations to
53:29build this model
53:30it's basically some fitting so we'll
53:32adjust those characteristics to match
53:35the switching uh pulse and then we'll
53:38adjust l1 here
53:39to match the lightning test and so then
53:42this
53:43surge arrester is valid for switching
53:44and lightning so that's the one
53:46recommended
53:47i really invite you to take a look at
53:48the video that we've done on server
53:50store modeling to do that
53:52and download in our exchange platform
53:54the add-on which does
53:56all that automatically yeah so this um
54:00this model you enter basically all the
54:02information you need
54:04actually for that you also need to know
54:05the number of columns in parallel and
54:08the length of the surge arrester okay so
54:11you enter all that in here
54:13and it will automatically create this
54:14model for you so
54:16fairly easy at the end of the day
54:23okay so here you have the way it's it's
54:25built if you want to do it yourself
54:29okay so this we discussed already the
54:30voltage source where you put a long
54:32transmission line to avoid reflection
54:40okay so let's sum up one more time uh
54:43what you have to do
54:44so the first thing will be to uplink the
54:46like apply lighting strike at default
54:48points to determine the critical
54:50current at um at each step
54:54when you try to find a critical current
54:57you compare the maximum voltage
54:59with the bil of each equipment
55:02but actually remember that we have to
55:04consider the bil with
55:06some safety coefficient safety factor
55:08amplitude
55:10correction factor and all that so what
55:12you're going to do is
55:14when you monitor the overvoltage instead
55:17of comparing with the bil you will
55:18compare to the bill divided by
55:20all these coefficients so safety
55:24coefficient
55:24safety factor atmospheric correction
55:27factor and altitude factor the
55:29atmospheric correction factor depends on
55:31the pollution mostly
55:33right so all that take a look at the
55:37you know for the ieee or iec standard to
55:40see what the value of those coefficients
55:41how to find them
55:42we had a lot of references in the origin
55:45and
55:46classification of all hbo
55:49and so then when you have this um
55:52when you have the critical card at each
55:54tower
55:56then you can determine the mean time
55:58between failure
56:01if you've watched the switching
56:03insulation combination video
56:05you've seen that we have two methods
56:07deterministic and statistical for
56:09non-self restoring insulation and
56:12self-resting insulation
56:14but like i said at the beginning of this
56:16one here there is no need
56:17um for this for lightning you can
56:21you can use the deterministic method for
56:23all type of insulator
56:25because of the low dispersion of
56:28the insulation correct the the waste and
56:32probably the western voltage probability
56:34okay so once
56:36you have uh your critical current
56:40you use here and a standard
56:43probability function to determine the
56:46probability to have
56:48such an intensity such a
56:52current constituent like me for example
56:55you remember i did here my simulation
56:58for the first hour and i found 75
57:02kilohertz for the critical current
57:06then i use this equation here
57:11it will tell me the probability to have
57:13such a lightning
57:14and i will assume it's the same
57:16probability
57:18for the entire distance covered by this
57:20tower so basically for half of
57:22from half of this line to half of this
57:24line i will assume this critical
57:26current just so i don't do so many
57:28simulations you have to simplify the
57:29problem
57:30and then you have this probability for
57:33all this section of line
57:35then you will do the same thing here and
57:36you will have the same probability there
57:43okay so once you have the probability
57:46like that for each section
57:48you can calculate the annual number of
57:52straw of
57:53strokes okay so for that
57:56the way you will calculate that is the
57:58ground density
57:59multiplied by multiplied by twice
58:03r a or a i will
58:06go back here to show you what it is
58:17go back where i was
58:22and b is the distance between each
58:24conductor
58:30okay so when you have so you have the
58:32annual number of strokes
58:34you multiply that by the probability you
58:37get
58:38for each section of line and
58:42k t it's an extra coefficient to add
58:46in the insulation coordination procedure
58:49which depends on the tower resistance
58:54okay so you will adjust this t 60 or 100
58:57percent for if it's a high resistance
58:58tower you use 100
59:00else to 60 of chance and so that's how
59:03you calculate now
59:04the number of backlash over for each
59:07section of line
59:08and you do that for expand
59:11until the critical distance and you can
59:14add up
59:14all these numbers to finally get
59:18uh the number of um the
59:22the annual number of backslash over
59:26okay so it's um it seems like uh it's
59:28difficult but it's actually not so
59:30difficult
59:32okay so if you need more readings
59:36you have here some references i
59:38recommend you to take a look at

About this transcript

This page contains the full transcript of Insulation coordination – Lightning transient overvoltages by EMTP, generated from the public captions YouTube serves with the video. The transcript has 8,518 words across 1,525 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.