Smart Meters and PV diversion

I have been running a PV diverter for several years now that I build from the circuits on this site. I’m very happy with the savings I’m making in heating my hot water. My electricity provider (Scottish Power) have now notified me of smart meter installations in my area. What I want to know is will my homemade diverter still work with a new smart meter? If not, can I just ignore the messages I’m getting, as I have no real need for a smart meter?

I’d like to hear some experiences for other people who are using a PV diverter with a smart meter.



The diverter will still “work”.

What happens next will depend on the size of the “energy packet” that your new meter uses (Learn | OpenEnergyMonitor). If the burst of energy that the diverter lets through each way is larger, then you’re going to be charged for importing even though the nett transfer of energy is zero. Unless you know the numbers for your new meter, there’s no way of knowing what the result will be.

It’s your decision, but if you’ve no desire for a smart meter, and you get the equivalent information from your own kit, I can’t see any advantage for you, and there’s a possible big disadvantage - you’ll effectively be paying the network to store your energy.

I too use an “energy bucket” style diverter for water heating, and I’m getting a smart meter, to take advantage of low tariffs to charge our EVs.
Is there any knowledge out there on the energy bucket size (if any) of smart meters?
Alternatively would it take a PWM style diverter (ie one that diverts in the microsecond range, not seconds) to solve this?

I believe - but with no firm data to prove it - that some must work down around the 1-cycle time frame. It should certainly be possible to do that, but impossible to deal accurately with reactive loads with an averaging period below 1 cycle.

I think you need to wait and see, and be ready to adjust the bucket size to see what effect it has.

I’ve often thought about designing a PWM power switch - and that’s as far as it’s got.

I’m not sure I understand what the difference is between the “Mk2” diverter and the “PLL” diverter, here at Learn | OpenEnergyMonitor - if indeed there is one, in practise.

Can anyone comment?

The difference is in the way the measurements are made. The diversion principle is the same.

Robin’s Mk2 runs the ADC interrupt-driven at the maximum sample rate, asynchronous to the mains. So the samples drift with each mains cycle, because there is not an exact number of samples per mains cycle. MartinR’s PLL phase-locks the ADC to the mains, so there is always the same exact number of sample sets always aligned exactly the same on each mains cycle. This necessitates rather fewer sample sets per cycle.

1 Like

I have been using Robin’s MK2 code and design now for many years and recently had a Smart meter installed so I could take advantage of cheap rate charging of my EV. I have now discovered that I am now clocking up usage when I wasn’t before the change over. I have gone through a re calibration of my diverter but to no avail. It is ‘better’ in NORMAL mode than ANTI_FLICKER but still clocks up usage.
I think this is possibly due to the ‘energy bucket’ of my smart meter (Aclara SGM1412-B) being non compatible with some of the default values in the code and I wondered if anyone has found out how to determine it’s value.
Paul Richards

The short answer is, we don’t know very much about the “energy bucket” - or indeed if there is one of any significant size.

Take a look at Robin’s website ( Only a day or two ago, he posted a new sketch “Basic PV Router sketch with faster control” aimed at alleviating some of the problems with smart meters.

Hopefully someone has learnt more regarding the OP’s question. My guess is that ‘smart meters’ will retain the energy bucket characteristic but the bucket size may vary with meter type and make.
Is there any new information?

Other than Robin’s experiences (and I understand he’s got a healthy market in a country where the meters are a lot more finicky than in the UK) there’s not been much reported - either successes or problems. I guess that’s a good sign.

I can recall just one comment recently that suggested that the energy packet was similar to (or even the same as) the meter it replaced.

It would appear that many/most smart meters do have the “energy bucket”, and that it’s at least comparable in operation to the analogue Ferraris meters.

Thank you Robert.