How might the proposed changes to UK mains voltage limits affect older electrical infrastructure?

As with lots of things the answer is it depends.  EG. 6amp rcbo trip time at 207v vs 230v vs 240v vs 253v 

The primary effect of different voltages relates to the current that can flow for a given fault or load (Ohm's Law: \(I=V/R\)). A lower voltage (207V) will result in a slightly lower current than a higher voltage (240V) for the exact same circuit resistance: For any given current level above 6A, the trip time is determined by the fixed trip curve (eg Type B or Type C).The only practical difference is that a specific fault resistance might draw a slightly lower current at 207V compared to 240V, potentially moving the fault to a different point on the time-current curve and resulting in a marginally different—perhaps slightly longer—trip time

The above being said a pump or a motor at the low end of the voltage spectrum might start to seem underpowered or run slower which could be an issue for some that have the pump set to a run time rather than a volume or sensor on and sensor off.

Caveat time.

240v in the UK became 230v nominal ages ago.  (about 1994) but in Yorkshire (other regions are available) 250v and above are quite regular.  The DNO/DSO/NG are probably in no rush to change the transformer tap to 207v but they reserve to use the right of 207v IF it is needed espcially for single phase.  This is probably being lead by the amount of re-newable energy be ProSumed/Generated

Its also worth noting that it would be a most odd situation where you saw the full variation from 253 off load to 207 on load, partly due to the heat losses in cables becoming more than 20% of that in  the load, and partly because you would have problems to arrange circuit protection safely. A consequence of  droop of 20% under  load is that the PSSC is only 5 times the full load current -  and given that most fuses will carry 100% overload i.e. 200% of nominal rating, for ages, and at a multiple of less than 5 times a B type breaker may not prompt trip and a C type is guaranteed by design, not to, then unusual measures (well earth fault relays and shunt trips) are needed. Also by the time you add on the permitted voltage drop in the final installation the socket in the garage at the end of the garden in the house at the end of the street main is looking very poorly ;-) 
You would only really expect to see both extremes when at some time of day local generation reverses the slope of the voltage drop, and then at other times heavy loads take the full quota.
I'd expect this concession to be used by the DNOs as a reason not to adjust things, rather than as a basis for re-setting transformers where no problems are reported, and perhaps allow slightly longer runs on new housing estates where in the past 2 transformers would have been used but were only just required.
Mike.

trip time at 207v vs 230v

Just to add to Mike's reply - in all this we're talking about voltages at two different parts of the system - the nominal supply voltages are as at the consumer's supply terminals (i.e. include the effect of voltage drop along the supply lines) whereas for loop impedance calculations the physics is really based on the voltage at the substation - the voltage drop along the supply lines during a fault is taken into account of by including Ze in the calculation. (A while back, ADS calculation took this into account, based on Uoc rather than Uo). It's very unlikely the substation would only be pumping out 207V as there would be no margin for v.d. in the supplier's lines.

   - Andy.

Would Voltage drop calculations needs to be looked at and possibly adjusted in BS7671?  As an example

8A current, 30m length, 18 mV/A/m value

18 * 8 * 30 over 1000

Volt drop 4.32 volts

Supply at 230 V: The voltage at the load would be 230v - 4.32v = 225.68v

Supply at 207 V: The voltage at the load would be 207v - 4.32v = 202.68v

4.32 is 1.88% of 230v

4.32 is 2.09% of 207v

Thus with the current 3% and 5% allowances for volt drop would the 5% limit need to be lowered in time or would product specification/requirements deal with it? 

Would Voltage drop calculations needs to be looked at and possibly adjusted in BS7671?  As an example

8A current, 30m length, 18 mV/A/m value

18 * 8 * 30 over 1000

Volt drop 4.32 volts

Supply at 230 V: The voltage at the load would be 230v - 4.32v = 225.68v

Supply at 207 V: The voltage at the load would be 207v - 4.32v = 202.68v

4.32 is 1.88% of 230v

4.32 is 2.09% of 207v

Thus with the current 3% and 5% allowances for volt drop would the 5% limit need to be lowered in time or would product specification/requirements deal with it? 

well its an interesting point, because meeting the product standards only requires the design to work  down to -10% of 230 - now at the moment we lose 5% of that between the substation and the meter, and the other 5% is available for the voltage drops in sub-mains and final circuits and so on. This way, the 'just passed 'CE marked product plugged in at the end of the extension lead at the far point of the installation is happy.
Ignore for now that a lot of private transformer sites that have a large extent, like factories hospitals and schools already take all of the 10% and sometimes more, for drops in their own wiring, and looking more at the effect on the man in the street.  Once the voltage at the meter is as low as 207, then actually no CE marked appliance is guaranteed to work anywhere other than very close to the meter indeed, as there is in fact no provision for a sensible amount of voltage drop in the wiring of the installation itself.

I suspect it won't be an issue except for a few fussy appliances, and not at all in small flats and bedsits, but in larger places with annexes and so on hot showers may be a bit cooler and toasters a bit slower than the worst case the makers are supposed to have tested when they apply the CE mark. 
Ideally we'd ask the product test spec lower limit to be set another 5 to 10% down to allow for realistic installation side voltage drops. However, I bet it won't happen any time soon, and 190V, might be realistic, but sounds a long way down.

Opening the 'acceptable' voltage eye too wide also rather degrades the idea of the O-PEN detection by voltage deviation as it decreases the range of fault conditions that can be detected with good confidence.

Mike.

mapj1 said:

This way, the 'just passed 'CE marked product plugged in at the end of the extension lead at the far point of the installation is happy.

Depends upon the length of the extension lead! I once could not get a chum's (table-top) bandsaw to start, but he told me that I just had to start it off by hand. It was daisy-chained at the end of two 50 m extension leads.

Surely, appliance manufacturers take into consideration voltage drop. Even then it is not much. Your 3 kW fan heater probably will not care, but perhaps a washing machine with a 2 kW heating element might. Let's make it a very old one with an ordinary 500 W motor; and let's put the laundry 20 m away from the DB. At 11 mV/A/m, that is about 2.4 V

So, we are talking about a voltage drop from the meter of the order of only 1%

except, like the band  saw. that 500W motor might take 10 amps or more on start up, and then the droop is excessive and perhaps it is very slow to start or perhaps it never does and just sits there and smokes...
Flats and small houses as you suggest with perhaps less than 20m of cable between meter and load are probably OK. But I am aware for example of odd cases of tumble dryer motors not lasting well (more precisely the start and run capacitors inside not lasting well ) in garages where the power supply is weedy (2.5mm SWA radial). Then there are places where the meter is at the opposite end of the drive to the building or there is a supply to the stables or something - the voltage drop allowance in the regs is there for a reason, and sometimes we do use all of it.

Appliance makers do take account of voltage drop, and currently test at 207 worst case for that. I'm noting that may now not always be 'low enough' to be confident with that appliance in all places it might be used.
Personally I think a better long term fix is to encourage the use of 3 phase to the house, and get rid of some of the neutral side voltage drop or rise that way.
Mike.