SVL is a new thing for me - so this is all guesswork (but might stimulate some useful discussion). The way I'm thinking - we know single core (non-ferrous) cable armours on a.c. circuits will develop a voltage between their ends (basically the armour forms a 1:1 transformer with the core, if a horribly inefficient one) - and generally we try to arrange things to keep this at safe levels (e.g. below 25V or 50V) and ideally try to avoid circulating currents in the armours. I think usual UK practice normally involves solid bonding of the armour at at least one point, with either gaps in the armour, or connections to armours from all phases 'rotated' at intermediate points so the induced voltage tends to cancel out.

Like I said, the SVL is new to me, but conceptually from what you describe it sounds a but like an SPD - shorting to earth if an overvoltage is present. I'm guessing it would be positioned where the armour would be normally gapped, intended to introduce additional earthing (or bonding?), but at the expense of allowing potentially large circulating currents. Perhaps when the cables are carrying substantially larger than the normal design current (e.g. during fault clearance times)?

So I'm guessing the thinking is that in normal circumstances you want to avoid circulating currents (as it's not only a waste of power, affects the voltage drop in the main circuit and generally overheats things - always bad for any conductor and especially its insulation) - so rate the SVL so it normally isn't called into action (i.e. > the voltage it would normally be exposed to) - but still lower enough that it can work on rare occasions and for short durations where the heating effect from circulating currents would be tolerable.

or I might be barking up completely the wrong tree....

    - Andy.

Andy and anyone else wondering what this is about - this intro to the topic may be helpful

https://www.powerandcables.com/svls-sheath-voltage-limiters-for-protection-of-mv-hv-cables/

Generally it is only needed when the currents within the armour of a cable do not cancel and induced sheath current would otherwise be a problem.

Cable with all three phases present inside the armour do not have this problem.

Mike

to determine a sheath voltage limiter (SVL) should have a continuous operating voltage of Uc higher than the calculated voltage at the cable end.

my point is why to choose higher than calculated voltage and not the same or less as to choose higher than the calculated value could be not safe for the cable system. 

It is important to realise that the intention of the SVL box is to ensure that the sheath/ armour is interrupted, being in effect broken into short sections so that the end to end induced voltage in any one section is non-hazardous (exactly how long that section is,  first requires the induced voltage to be calculated, and secondly a decision  on what non-hazarous means in this context).  The SVL is then in its high resistance state most of the time, and in normal operation no sheath/armour current flows.
However, during a fault, when we do need current to flow in the sheath/armour of the cable, we require the SVL, or SVLs to  breakdown into their low impedance condition, and carry the highest credible current, for long enough for whatever protective device is involved to operate and remove the supply, ideally with some margin.

The SVL is usually a large block of semiconductor, but if it helps you can imagine it behaving in some ways like a spark gap in that it has a definite 'flash over' threshold,  and the ability to carry high currents for a short time, and then a recovery time once the current has dropped, for it to go back to the high impedance state,

This is similar to a surge protector, but the way it is tested and specified is a bit different.

What we do not want is the SVL to 'flash over' during normal operations, like brief overloads, switching transients and so on, so the threshold voltage and the segment length need to be set so that the operating voltage is perhaps twice the maximum expected  armour voltage.

Mike

For a description of induced sheath voltage

https://electrical.theiet.org/media/1678/voltages-induced-in-the-non-magnetic-metallic-sheaths-and-or-armour-of-single-core-cables.pdf

For the maths that allow you to calculate it

https://mycableengineering.com/knowledge-base/sheath-induced-voltage-circulating-current