EVs, Street furniture, PME and TT configurations

The floating secondary problem is very real, and even on small transformers used for charging portable kit the effect can be, as you say "tingly".

Before the  mandated efficiency targets drove everything towards switch mode, (where the frequencies are higher, but the transformer is smaller, and now the leakage is not at 50Hz), the solution used to be the split bobbin mains transformer.
https://www.belfuse.com/product-detail/signal-transformer-high-isolation-split-bobbin-low-power-transformers

The problem is that these are fine for a few hundred watts, but the core, and therefore losses, end up quite a bit larger than if you allow the windings to interleave, and a toroid would be smaller still. However (of course) if you do that you bring the 'plates' of the coupling capacitance closer and so the interwinding capacitance gets larger, and is anyway more noticeable on on bigger transformers.

It is possible to interrupt the capacitance and have an interwinding screen and either earth it, or in some odd special EMC related cases, connect it to another voltage that is not actually earth but does not have such a high impressed alternating voltage as the primary winding. After earthing the most common thing is to neutral it, followed by connecting it to a mains derived DC bus. In such a case the screen must then be insulated from the output windings to a similar degree as if it were another mains winding.

Transformers like this are not hard to make, but are not the standard item, and this is reflected in the price at least for quantities where it is not worth tooling up a production line, and they are individually made.


In the car situation, you are not too worried about this, as you would earth one side of the secondary anyway, but the bigger problem is that it is very little help, as while you side step the local PME you are still mixing earthing types if there is other class 1 kit using electricity nearby. It is no better than running an SNE cable back to the substation and saying 'this supply for the charger is TNS' Which is great, if it is in the middle of an otherwise deserted  car park, but no help at all if there are PME supplied items within reach of the car  that will be plugged in. And that may be another issue - as the car plus cable may be a few metres long and could in some cases will be parked on one side or the other, the exclusion area for simultaneously accessible ought if anything to be bigger, to include the area of the vehicle .


The other question is if the risk is credible. There is a price to safety, and when it comes to vehicles  they are inherently a lot higher risk than electricity. It may well be that all this is disproportionate, and if we wanted to save more lives we'd ban the vehicles all together.

Normally to see if an activity is worthwhile you assign a cost to a life lost which may be a few million pounds, then divide the cost of the extra precaution times the risk.

For simple sums if my life cost is 1 million, and there is a new piece of safety kit that costs £10, then it  is only worth installing them everywhere if the risk it mitigates is more than 1 in 100,000. Now the hard part is always getting the probabilities and full costs right and for that it stands or falls by the accuracy of available data. It is hard enough to be sure if RCDs are a sensible idea, given the low numbers of electrical accidents both before and after they became commonplace, let alone with something  we have so few of at present. Once we get a decent number of accidents we will be in a better place to understand the risk quantitatively, rather than just emotionally.

The floating secondary problem is very real, and even on small transformers used for charging portable kit the effect can be, as you say "tingly".

Before the  mandated efficiency targets drove everything towards switch mode, (where the frequencies are higher, but the transformer is smaller, and now the leakage is not at 50Hz), the solution used to be the split bobbin mains transformer.
https://www.belfuse.com/product-detail/signal-transformer-high-isolation-split-bobbin-low-power-transformers

The problem is that these are fine for a few hundred watts, but the core, and therefore losses, end up quite a bit larger than if you allow the windings to interleave, and a toroid would be smaller still. However (of course) if you do that you bring the 'plates' of the coupling capacitance closer and so the interwinding capacitance gets larger, and is anyway more noticeable on on bigger transformers.

It is possible to interrupt the capacitance and have an interwinding screen and either earth it, or in some odd special EMC related cases, connect it to another voltage that is not actually earth but does not have such a high impressed alternating voltage as the primary winding. After earthing the most common thing is to neutral it, followed by connecting it to a mains derived DC bus. In such a case the screen must then be insulated from the output windings to a similar degree as if it were another mains winding.

Transformers like this are not hard to make, but are not the standard item, and this is reflected in the price at least for quantities where it is not worth tooling up a production line, and they are individually made.


In the car situation, you are not too worried about this, as you would earth one side of the secondary anyway, but the bigger problem is that it is very little help, as while you side step the local PME you are still mixing earthing types if there is other class 1 kit using electricity nearby. It is no better than running an SNE cable back to the substation and saying 'this supply for the charger is TNS' Which is great, if it is in the middle of an otherwise deserted  car park, but no help at all if there are PME supplied items within reach of the car  that will be plugged in. And that may be another issue - as the car plus cable may be a few metres long and could in some cases will be parked on one side or the other, the exclusion area for simultaneously accessible ought if anything to be bigger, to include the area of the vehicle .


The other question is if the risk is credible. There is a price to safety, and when it comes to vehicles  they are inherently a lot higher risk than electricity. It may well be that all this is disproportionate, and if we wanted to save more lives we'd ban the vehicles all together.

Normally to see if an activity is worthwhile you assign a cost to a life lost which may be a few million pounds, then divide the cost of the extra precaution times the risk.

For simple sums if my life cost is 1 million, and there is a new piece of safety kit that costs £10, then it  is only worth installing them everywhere if the risk it mitigates is more than 1 in 100,000. Now the hard part is always getting the probabilities and full costs right and for that it stands or falls by the accuracy of available data. It is hard enough to be sure if RCDs are a sensible idea, given the low numbers of electrical accidents both before and after they became commonplace, let alone with something  we have so few of at present. Once we get a decent number of accidents we will be in a better place to understand the risk quantitatively, rather than just emotionally.