Sorry, no idea about G59 relays.

Posted By: ActivePassiveI am assuming for the moment a 17 kW G59/3 connection is feasible given my incoming fuse is 100A (24kW), so the cable at least should be able to support 10.5 kW of generation?

That doesn't follow. It's not (just) the current carrying capacity of the cable but also a matter of keeping the voltage at each of the consumers in the legal range.

The ESQR [¹] says that normally mains voltage as supplied to the premises should be 230 V +10%/-6% so 216.2 to 253 volts. In practice they seem to normally aim for about 240 V (the old standard up until 1994) presumably because a) it's what's already there so they won't get complaints if it doesn't change, b) they sell more energy for resistive devices (like most heaters), and c) it gives them a goodly margin for voltage sag on the days when all the consumers on that phase gang up to draw their full 100 amps.

So there's a 23.8 volt margin for voltage sag but only 13 volts of headroom for over-voltage and therefore about half the current capacity in the opposite direction (e.g., on a sunny day when all your neighbours have had their 3.84 kW installed but gone out for the day leaving nothing switched on). And maybe there are also considerations of transforms being designed to work one way - in principle transformers work both ways but I think that in practice they're less efficient going the other way.

Bear in mind that the DNO has to consider the case that everybody on the line eventually gets 16 A/phase PV on your line. If they don't charge you for any required upgrades now they might well have to pay for it themselves when the last person gets theirs installed and everybody's inverters overvoltage trip.

[¹] The Electricity Safety, Quality and Continuity Regulations 2002
2002 No. 2665 PART VII Regulation 27
http://www.legislation.gov.uk/uksi/2002/2665/regulation/27/made

Ed, Barney thanks for your comments.

Ed, do you know how they calculate the over voltage margin based on the max kW of the generators versus number of consumers (houses) on the same phase (on the 240V, house side of the transformer)? For example if the overnight supply voltage was 240V and every home had a 3.84kWp solar PV system facing south, what over-voltage would you get on a very sunny mid-day?

When I spoke to them on the phone they talked about 'reinforcement using 300 wavecon' but didn't explain why. Would this be a way of reducing the nominal resistance of the cable?

Given their initial estimate is £20K for reinforcement its something I would want to avoid. If this is the case switching to a different phase or bringing 3 phase on site might be the cheaper option? The existing overhead cable looks pretty meaty from the ground and all the local cabling was recabled about 5 years ago.

I thought I would try a little experiment - turn everything on in the house and see what impact it has on the voltage and then potentially imply what impact in the opposite direction a solar array might have?

The voltage in the area as monitored by a plug-in appliance monitor is quite high - 235V to 240V during the day and 245V in the middle of the night. I switched everything on - the induction hob, oven, microwave, electric heater, HW immersion, kettle drawing about 16kW according to my Owl (not accurate if inductive or capacitive load) but it seemed about right for the appliances I switched on.

The voltage at a mains socket in the kitchen dropped from 240V to 210V. A socket in the garage where the meter and incoming cable is, dropped from 240V to 225V.

It does suggest there is some issue with load/generation in my house if 15 kW drops the incoming mains voltage by 15V - my gut feel is that this voltage drop is quite high, or is this normal? The experiment might be saying more about my connection (10m) from the garage underground to the telegraph pole than the onward connection to the transformer?

If you assume symmetry between load and generation and the voltage on a sunny summers day could be as high as 245V, then you would have 8V leeway to the max 253V allowed, which approximates to about 7 kW of generation at the incoming supply?

I would appreciate your thoughts on this experiment and whether it has any validity and/or whether anyone else might have tried this in a more urban environment rather than my more rural setting?

What I also don't know is what impact this load had on the voltage difference between neutral and earth which I have seen as high as 35V in the house and whether the appliance monitor measures voltage to neutral (I presume it does) or to earth? The house has an earth spike which I believe is common in rural areas with overhead telegraph pole supplies?