Dear colleagues, let me please start a series of “green posts” on electromagnetic transients and protection against them. In general, I have to admit that there are too many experts in this area in the world, and it seems to me sometimes, they are ready to fight brutally for their right to exist under this sun, suppressing any dissent and competition. On this occasion, I can say right away that I’m not going to fight. I see my task only in sharing educational content. When a professor is asked the same questions year by year, he is not interested in answering them any more (or taking money for it), and he simply publishes a tutorial. This is how the education system works.

Today let us assume that we have a dead-end circuit where a 500 kV overhead line (OHL) feeds a 500 kV switchgear having, say, one instrumental voltage transformer (IVT, of a capacitive type) and one power autotransformer (AT). Let us assume that as a result of lightning activity, lightning waves form on the line, which propagate to the switchgear and create transients there. Let us try to understand what will change if we move the surge arrester (SA) between the IVT and AT.

We will not do statistical calculations, but simply consider that a wave (with a peak of 2500 kV and a front of 1 µs) propagates to the switchgear (at higher voltage, the OHL’s insulation will overlap). We will assume that the distance from the IVT to the AT is 100 meters, and we will consider two options:
1️⃣ SA is located at an equal distance from IVT and AT.
2️⃣ SA is located at AT (as the most expensive equipment).

We will vary the SA’s residual voltage (Ures, at a 8/20 µs current of 10 kA): 900 kV, 1050 kV, 1200 kV. It should also be recalled that the peak value of a lightning impulse, which is permissible for 500 kV equipment insulation, is usually about 1500-1600 kV.

The upper oscillograms show that, despite the equal distances L1=L2:
✅ peaks on the IVT and AT are not equal to each other;
✅ peak for IVT is almost independent of SA, whereas for AT the peak is proportional to Ures of the SA.

The lower oscillograms show that at L1>>L2:
✅ AT is better protected;
✅ processes on IVT have almost not changed.

There could be a conclusion: in addition to the distance from the SA to the equipment being protected, it is very important how this equipment is positioned relative to the SA along the way of the incoming lightning wave:
➡️ equipment is located before SA (case of IVT, circuit breaker, etc.);
➡️ equipment is located after SA (case of AT).

No matter how well we install the SA at the power transformer, the equipment connected to the OHL will probably be insufficiently protected, and therefore it is better to install a second set of SA at the line input to the switchgear.

You can find these and other things, including detailed statistical protection calculations, in the books on my website.