i have modeled a lightning strike as below(current source),and i want to impose it to my HVDC line. im using 1.2/50 microSec model of lightning but i need to know the current peak. ill be glad if any one could help about peak value of lightning current in mentioned situation.
Actually 1.2/50 microseconds is waveform of voltage impulse, not current impulse. Waveform of lightning current impulses, for example, for surge arrester testing is 8/20 microseconds. (30/60 or 45/90 microseconds for switching current impulses). Peak values for lightning current impulses: 1.5...20.0 kA; for switching current impulses: 0.25...2.0 kA.
As far as I know, the lighting strikes are usually modeled by time-dependent voltage function of a double exponential waveform with a peak that may reach 15 MV. Therefore, the consequent current flow depends on the impedance of the strike path. Consequently, you need to simulate the strike voltage waveform, then apply it to the RLC circuit and the surge protection model of the HVDC line to see the current flow and the overall response of the circuit.
According the relevant literature, a double-exponential function don’t describes correctly the shape of lightning current. For testing purpose (of surge arresters, as example) it was adopted double-exponential current shapes with different time parameters in order to simulate different stages of lightning and this parameters are specified in dedicated standards.
But, nowadays, the lightning current is considered to be describe adequately by Heidler function (you can find many papers about this issue). Regarding the amplitudes: all lightning parameters (peak values included) obey of statistical distributions (approximately of log-normal shape) with different values for positive lightning (10…20 % from total such events) or negative lightning (80…90% from total). So, you can choose either the values corresponding to 50% probability to be exceeded, either the values corresponding to lower probability (as example 5% or 1%, which means rare occurring events). Such values could be found also in literature. As example, the peak values of currents which can be exceeded with 50% probability are: 31 kA for negative lightning (first stroke) and 35 kA for positive lightning. The peak values which can be exceeded with only 5% probability are: 69 kA for negative lightning and 250 kA for positive ones. Generally, as higher object is, the higher the lightning's current value is. The time parameters can be found also in relevant literature. As example, you can find information in "Electra", vol. 69 (1980) “Lightning parameters for engineering application”, CIGRE Brochure no. 063 (1991) etc. The choosing of data must be according your real goal.
I think Marian gave a rather detailed answer. Additionally, we can say that for a specific lightning strike location, there are isokeraunic maps showing the number of storm days per year per km2. Another phenomenon, coupled with lightning strikes, is represented by multiple lightning strikes. Multiple lightning strikes differ from the first lightning strike (eg amplitude, steepness) and are also statistically processed in the relevant literature.
I think Marian's answer is a good one.
One thing maybe to add. In laboratory you connot reproduce real lightning surges, therefore you are tied to double exponential impulses, just like the 10/350 or as mentioned above 8/20 (and various other impulses, mainly consisting of two time constants and a peak value).
When simulations are done, there are some mathmatical expressions, e.g. Heidlerfunction, which is really favorble (because it is three times continuously differentiable), but basically it is also a double exponential function with a slow (due to the exponent, which is fixed to 10 for the IEC 62305). For a more realistic representation CIGRE has proposed it's own function, that can use steepness, rise time, peak value and tail time independetly from each other.
When you only want to simulate the event, try to use ATP/EMTP (also a Spice Simulator). Hereby you will find both lightning sources already implemented.
Kind regards
You could review this paper, because the peak current has a dependence with latitude. See section III.A. Regards.
If your idea is to simulate lightning surges in HVDC, maybe the answer to your problem is to use Monte-Karlo simulation in which by generating random values of the lightning peak you will be able to simulate the consequences of the surges. For modeling the lightning surge you should use Umax[exp(-at) - exp(-bt)] (voltage-time depended source)
There is a range of currents of lightning strokes that terminate to phase conductor of the transmission line - not intercepted by shield wire. You can calculate the maximum shielding failure current based on the attached article (Section 2). Typical values are 15 kA - 30 kA.
If the lighting terminates to shield wire or un-shielded phase conductors of transmission lines, then you should consider the lighting peak current distribution (see CIGRE WG C4.407, or paper of J. Montana above) - depends on the probability that you want to take into account the selected peak current. Typical values are 100kA - 200kA in similar studies.
With regards to current waveform you shall not use 1.2/50us but a current waveform that depends on the polarity and the peak current of lightning as determined by CIGRE WG C4.407. Typical value for negative lightning 30kA: 5.5/75 us.
We use intensively lighting detection system for years. For each year, we have statistical analyses about many parameters for lightning strikes. Current values for lightning strikes have wide range and frequency. Small current can be in a range of 2 to 3 kA and maximum we record 250 kA. But the most common value on our geographical latitude is around 7-8 kA till 20-25 kA.
- Badges
- Science topic