Is a new electrical ecosystem needed and what could it offer?

I'm not sure about the need for this.


Also looking for gaps in that stair case I'd argue the other way, at least in the western world, 230V single phase will soon not be enough as we try to make rechargeable lorries etc,

If we want anything it is probably a layer above our common 230/400 and less common 400/690 single/three phase, perhaps adding a tier of 690/1k2  for longer distances, and better compatibility with electric vehicle voltages that need to be in the high hundreds to kV region, and modern solar panels where the string voltages are also a few hundred.


As a crude rule of thumb, for distributing significant currents  without needing to oversize cables relative to thermal ratings to mitigate excess voltage drop, one meter per volt is a good starting point. so a 10KV line may go no more than a small multiple of 10km before cable loss means it is worth having transformers at each end,  and in the same way 12V DC may go 10-20m and run out of steam.

This is based on copper or aluminium cables and 5% to 10% voltage drops. If the interest is high efficiency, and for this it should be than really you do not want to lose more than that in the cables. Equally if you want to use steel wires for reasons of economy, these are more resistive, and you may want to go the other way and use slightly higher voltages for a given distance - the SWER systems in parts of Africa deliberately have a higher voltage to distance ratio, to allow the use of what is in effect strands of galvanised steel fence wire as the single HV conductor. (though these systems are not without their own problems)


At 48v generation and load will need to be within a few tens of metres (perhaps up to 100m) of each other, so there is little scope for spatial averaging by interconnection between islands.


The 48V Telecom system is not such a great fit to that model , as at both ends clever electronics is working hard to change it into something else - in the handset circuitry that mostly runs in 3.3V and 5V CMOS has to step up the last bit to pretend to be a moving iron earpiece and carbon mike, and at the exchange, the carefully spoofed same voltages are stepped down again for digitisation.Of course it is also a signalling  system not power transmission, and  so the length to voltage consideration does not really apply. (and to show there is no real voltage distance link for signalling, for gigabit Ethernet, voltages of order 2Vp-p  go a few hundred metres on a good day.)


I am not familiar with the 48V cooker, I imagine there is a serious fire risk from the sort of currents needed to supply a few households cooking, and arranging reliable ADS on DC is not easy - welder like arcs can be drawn without the zero-crossing to aid extinction.


I suggest that more technical meat is needed to make a convincing argument - a comparably safe system in terms of shock risk, if the assumption is that there will be bare conductors, that would be far easier to integrate with grid equipment as well, might be something closer to the UK RLV 55-0-55 supplies used on building sites.

Mike.



edited 11/3/2021 to tidy up the explanation and assumptions of the voltage drop range limits.

I'm not sure about the need for this.


Also looking for gaps in that stair case I'd argue the other way, at least in the western world, 230V single phase will soon not be enough as we try to make rechargeable lorries etc,

If we want anything it is probably a layer above our common 230/400 and less common 400/690 single/three phase, perhaps adding a tier of 690/1k2  for longer distances, and better compatibility with electric vehicle voltages that need to be in the high hundreds to kV region, and modern solar panels where the string voltages are also a few hundred.


As a crude rule of thumb, for distributing significant currents  without needing to oversize cables relative to thermal ratings to mitigate excess voltage drop, one meter per volt is a good starting point. so a 10KV line may go no more than a small multiple of 10km before cable loss means it is worth having transformers at each end,  and in the same way 12V DC may go 10-20m and run out of steam.

This is based on copper or aluminium cables and 5% to 10% voltage drops. If the interest is high efficiency, and for this it should be than really you do not want to lose more than that in the cables. Equally if you want to use steel wires for reasons of economy, these are more resistive, and you may want to go the other way and use slightly higher voltages for a given distance - the SWER systems in parts of Africa deliberately have a higher voltage to distance ratio, to allow the use of what is in effect strands of galvanised steel fence wire as the single HV conductor. (though these systems are not without their own problems)


At 48v generation and load will need to be within a few tens of metres (perhaps up to 100m) of each other, so there is little scope for spatial averaging by interconnection between islands.


The 48V Telecom system is not such a great fit to that model , as at both ends clever electronics is working hard to change it into something else - in the handset circuitry that mostly runs in 3.3V and 5V CMOS has to step up the last bit to pretend to be a moving iron earpiece and carbon mike, and at the exchange, the carefully spoofed same voltages are stepped down again for digitisation.Of course it is also a signalling  system not power transmission, and  so the length to voltage consideration does not really apply. (and to show there is no real voltage distance link for signalling, for gigabit Ethernet, voltages of order 2Vp-p  go a few hundred metres on a good day.)


I am not familiar with the 48V cooker, I imagine there is a serious fire risk from the sort of currents needed to supply a few households cooking, and arranging reliable ADS on DC is not easy - welder like arcs can be drawn without the zero-crossing to aid extinction.


I suggest that more technical meat is needed to make a convincing argument - a comparably safe system in terms of shock risk, if the assumption is that there will be bare conductors, that would be far easier to integrate with grid equipment as well, might be something closer to the UK RLV 55-0-55 supplies used on building sites.

Mike.



edited 11/3/2021 to tidy up the explanation and assumptions of the voltage drop range limits.