Immersion Heater - PV electricity

[echase]

Did echase manage to get his design commercially developed?

Yes. My non-burst fire unit meets EN 61000-6-3 Electromagnetic compatibility (EMC). Generic standards. Emission standard for residential, commercial and light-industrial environments 2007, and BS EN 61000-3-3:2008 and EN 61000-3-11 for flicker.

I suspect the original EMMA device did too although I believe their more recent devices use high frequency pulse width modulation, which reduces the low frequency end but increases the high frequency so you need a screened cable to the immersion.

 

[echase]

ah, hold on they're different regulations, it's BS EN 61000-3-3:2008 and EN 61000-3-11 that are the regulations in question.

this diagram explains why the Immersun can't meet those regulations, as it shouldn't be able to cause a volt swing each time it cuts in of more than just under 0.3% or around 0.7V, but the location we had it at I measured around a 2V drop from switching on a 3kW load, so effectively every time the burst fire kicks in on each cycle it's causing around a 2V drop on that cycle.

Presumably this is why dishwashers and washing machines are restricted to around a 1kW element.

Design for EMC Part 6: ESD, dips and dropouts, etc.

http://www.reo.co.uk/files/handbook_en_61000-3-3_and_en_61000-3-11.pdf

I've been meaning to get back to 4-Eco about this tbh.

Excellent summary of the problem. I have built a couple of experimental burst fire controllers and installed them in 2 different houses. In both there was a significant level of lighting flicker, including neighbour’s lights, so I had to rapidly disconnect them. These were not houses at the end of long wires to the substations.

As I understand it you might be able to slow down the flicker to a frequency that is to the left of that graph and hence just about meet the standard at 3kW but it will be so slow that it won’t spoof the electricity meter into recording a zero export/import in all circumstances and certainly not when new/smart meters come in with the likelihood that they will be more sensitive to catching short bursts of power.

Also if you go too slow you may meet the standard but your neighbours will still see their lights flickering fast enough to be annoying.

The trouble with EMC etc. standards are that they are quite severe so it may be that you have failed even if you don’t see the flicker. Some houses might have impedances low enough so that no one notices the flicker, especially if you are not looking for it, but actually be producing voltage dips more that that graph allows.

One has to ask the question “Why do cooker and dishwasher manufacturers with all their research resources limit their burst fire controllers to 600-1000W?” It is because they can’t comply with flicker specs at higher powers. As they can’t what chance have we?

 

[suntrap]

Agreed. The only point I would take issue with is if newer meters would in fact be more "sensitive" to small reverse power flows, given that I don't see why there would be a technological limitation on the ability of current meters to detect this. I would have thought that installing a meter that behaved differently for no good reason would be unfair. Is the curious way the current ones work an attempt to make them behave more like analogue meters? I don't know - pure conjecture!

 

[echase]

In a not very scientific test I measured a 2 volt drop on my house with a DMM by turning on a 3kW kettle on and measuring the change in voltage at the consumer unit. I am only 200m and 16 houses from the substation. Using that figure (= 0.9%) and the above graph it shows that the switching rate has to be less than about 100 changes per min or 1.6Hz. That is probably too slow to definitely spoof a meter and has to be much slower at non 50/50 mark space ratios (half power). My friend’s was running at a “legal” 1Hz when the neighbour complained. Would stress though that this is in no way taking electricity one has not paid for, as it is returned again.

But before outright saying all faster burst fire 3kW devices fail it has to be said the proof is in an actual flicker test in a lab, as that graph is only an interpretation of the flicker standard from http://www.reo.co.uk/files/handbook_en_61000-3-3_and_en_61000-3-11.pdf and probably only applicable to 50/50 mark space ratios, which for a solar diverter is rarely the operating point. Someone may yet prove us wrong by managing to meet the spec whilst also stemming the neighbours’ complaints.