Beware SMETS2 meters

[mikemillar]

Some of you will have between following my thread at
Gateway 2 and Smart Meter accuracy
regarding over-reading by my new SMETS2 Smart Meter in conjunction with my Powerwall 2. A 'discussion' with EDF has been ongoing since November. However, it has now taken on a new lease of life - possibly because an EDF manager with an experimental home battery has been experiencing the same problem as me!

Tesla Energy have already been investigating but I have not feedback yet. However, I was telephoned by the EDF manager in quesition yesterday and as a result, at his request, have just sent a lenghty technical report with a lot of sample data for the EDF 'Blue Lab' technical people to chew over. The implication, he has told me, is that the new SMETS2 meters are designed to be capable of metering Supply Side, or 'Demand' Response energy trading - you get paid to have backup power on standby in case of a major frequency problem on the grid. This is limited to big players at the moment, but in the future, networks of connected home batteries and electric cars could participate, and metering will be required.

The current problem appears to be that the Powerwall 2 takes and returns small amounts of energy to maintain frequency parity with the grid frequency. Unfortunately it seems that they have not considered the effect of this on SMERTS2 meters and mine, at least is accumulating more than double this low level activity. Octopus Energy have mentioned to me in passing that they 'tend to install SMETS1 meters where they are supplying Powerwalls. One Powerwall owner I am in correspondence with on th4e PVOutput site has told me that he has dumb SMETS1 meter and it only records about 3 watt-hours on Peak rate compared to my 80 watt-hours (£200 per year for me!).

So until investigations are complete and a fix has been worked out, my advice is - Consider very carefully before you agree to having a SMETS2 Smart Meter installed if you also have a home battery.

There could be some red faces over this, politically as well as technically, but no doubt the politicians who are pushing the Smart Meter 'revolution' will find a way top wriggle out of it as usual, if the coronavirus doesn't get them first!

 

[Nhallwood]

Some of you will have between following my thread at
Gateway 2 and Smart Meter accuracy
regarding over-reading by my new SMETS2 Smart Meter in conjunction with my Powerwall 2. A 'discussion' with EDF has been ongoing since November. However, it has now taken on a new lease of life - possibly because an EDF manager with an experimental home battery has been experiencing the same problem as me!

Tesla Energy have already been investigating but I have not feedback yet. However, I was telephoned by the EDF manager in quesition yesterday and as a result, at his request, have just sent a lenghty technical report with a lot of sample data for the EDF 'Blue Lab' technical people to chew over. The implication, he has told me, is that the new SMETS2 meters are designed to be capable of metering Supply Side, or 'Demand' Response energy trading - you get paid to have backup power on standby in case of a major frequency problem on the grid. This is limited to big players at the moment, but in the future, networks of connected home batteries and electric cars could participate, and metering will be required.

The current problem appears to be that the Powerwall 2 takes and returns small amounts of energy to maintain frequency parity with the grid frequency. Unfortunately it seems that they have not considered the effect of this on SMERTS2 meters and mine, at least is accumulating more than double this low level activity. Octopus Energy have mentioned to me in passing that they 'tend to install SMETS1 meters where they are supplying Powerwalls. One Powerwall owner I am in correspondence with on th4e PVOutput site has told me that he has dumb SMETS1 meter and it only records about 3 watt-hours on Peak rate compared to my 80 watt-hours (£200 per year for me!).

So until investigations are complete and a fix has been worked out, my advice is - Consider very carefully before you agree to having a SMETS2 Smart Meter installed if you also have a home battery.

There could be some red faces over this, politically as well as technically, but no doubt the politicians who are pushing the Smart Meter 'revolution' will find a way top wriggle out of it as usual, if the coronavirus doesn't get them first!

This is interesting. I've just had SMETS2 meters installed 2 weeks ago as part of the bulb rollout. I've had my Powerwall2 since November. I noticed on the "old" meter (dual-rate) that I accumulated small usage of normal rate grid power even though I was powered throughout by the Powerwall and there were some noticeable little blips of grid activity, in particular during the night but also during the day Since the SMETS2 meters were installed I've used 6 units of normal rate electricity although probably about half of those are actual usage due to my car charging and the home usage going above 5kW during the day. If I extend the 3kWh in 2 weeks for the whole year then that would equate to 75kWh which at my current cost would be around £13 so annoying but a relatively small price to pay compared with the savings.

 

[charlesj]

Interesting even if it doesn't apply here as far as I know.

 

[JohnRatsey]

I think my meter is SMETS1 (installed several years ago) and I'm seeing at least 0.5 kWh daily drain from the grid even when household demand is satisfied by solar and/or battery. Here's a recent consumption graph from the meter data:

I'm not going lose sleep over that consumption as the cost of this electricity is less than half the daily standing charge (25p/day for Octopus Go tariff).

 

[sashton]

The general "leak" of about 60w average into a PW2/BUG2 system as measured by my Octopus SMETS2 meter but not seen by my Fronius MID certified meter is irritating. It probably costs me about an additional £75/year.
In addition to that, for about four months of the year I import additional power from the grid during off-peak. Unfortunately the SMETS2 billing periods are not correctly clock synchronised, averaging 2.5 minutes late. This results in an additional 0.42 kWh on average being incorrectly billed at peak rate per day. This results in an additional £5/year billed.
@mikemillar Do let me know how you get on as I too have an open case with Octopus over this (opened six months ago

 

[Solar1920]

I've recently had a SMETS2 meter installed and am starting to have the same concerns as you.
See below the reading from my smart meter from a day where the Powerwall supplied power for the whole day...

Below is the powerwall import details from the same day...

and Import details below from the Solaredge inverter for the same day. (the import is correct, but consumption is not correct as it is not setup to be aware of my other array and the powerwall)....

The powerwall and Solaredge have readings fairly similar around 0.3-0.4kWh, if these readings are to be believed it means i'm being charged for approx 1kWh a day i don't use, my consumption has never been below 1.4kWh a day since i've had the meter installed.

@mikemillar does this tie in with the sort of readings you are seeing?

 

[mikemillar]

This is interesting. I've just had SMETS2 meters installed 2 weeks ago as part of the bulb rollout. I've had my Powerwall2 since November. I noticed on the "old" meter (dual-rate) that I accumulated small usage of normal rate grid power even though I was powered throughout by the Powerwall and there were some noticeable little blips of grid activity, in particular during the night but also during the day Since the SMETS2 meters were installed I've used 6 units of normal rate electricity although probably about half of those are actual usage due to my car charging and the home usage going above 5kW during the day. If I extend the 3kWh in 2 weeks for the whole year then that would equate to 75kWh which at my current cost would be around £13 so annoying but a relatively small price to pay compared with the savings.

If you've used only 3 units over 14 days, with the other 3 accounted for, That's about what it should be. Unless the meter is set to record export as well as Import we are going to be charged a small amount that we're not really using (or at least putting back). I'm being paid FIT for 50% solar export and hardly exporting any - you can't have your cake and eat it (unless you are Boris....)

 

[mikemillar]

I'm not going lose sleep over that consumption as the cost of this electricity is less than half the daily standing charge (25p/day for Octopus Go tariff).

On my EDF 25p/5p split my 2kWh per day amounts to nearly £200 per year or 25% of my total bill - so I think I'm justified in loosing sleep over it!

 

[mikemillar]

The general "leak" of about 60w average into a PW2/BUG2 system as measured by my Octopus SMETS2 meter but not seen by my Fronius MID certified meter is irritating. It probably costs me about an additional £75/year.
In addition to that, for about four months of the year I import additional power from the grid during off-peak. Unfortunately the SMETS2 billing periods are not correctly clock synchronised, averaging 2.5 minutes late. This results in an additional 0.42 kWh on average being incorrectly billed at peak rate per day. This results in an additional £5/year billed.
@mikemillar Do let me know how you get on as I too have an open case with Octopus over this (opened six months ago

I'll keep you updated, @sashton. I've had a short correspondence with Octopus and they implied that they don't normally install SMETS2 meters in conjunction with Powerwalls at the moment. I think we need to wait for the outcome of my tech report and EDF's technical investigation. But I'll make the results available to those of you who have noticed problems.

 

[mikemillar]

I've recently had a SMETS2 meter installed and am starting to have the same concerns as you.......
does this tie in with the sort of readings you are seeing?

Pretty much, @Solar1920, although your average 30W/h is rather less than my 80W/h import when the Powerwall is supplying all the demand. However, the Tesla Spec is for a max 500Wh/day, which is <20W average. So your's is still too high. You seem use very little electricity - 1.4kWh per day? In your case although the excess demand metered is quite small it still amounts to a significant part of your total bill, and such small discrepancies are quite difficult to detect.

I think there may be a major incompatibility between the SMETS2 meters and home battery 'activity' that hasn't been properly thought through.

 

[McWill]

I can't say I'm seeing the same, attached are 2 images, the first a section of our SMETS2 30 minute readings and the second the recordings we made of the same period from the powerwall local api.

The meter readings for the three 30min readings starting at 20:30 all show a minimal 0.003kWh recorded which would be less than 0.2kWh for the whole day.

I don't know if it's significant but we have a Tesla Gateway 2 and a Landis Gyr E470 Type 5424 meter.

 

[Solar1920]

I also have a gateway 2 and a Landis Gyr E470 Type 5424 meter

I'm raising a support case with my supplier to see what they have to say.

 

[mikemillar]

I don't know if it's significant but we have a Tesla Gateway 2 and a Landis Gyr E470 Type 5424 meter.

View attachment 521965 View attachment 521966

That's exactly the same setup and meter that I have. I assume you have solar as well. You are getting the sort of readings I would expect, about 3Wh, at least when your home power demand is less than 5kW, between 20:30 and 21:30. @pgalbavy on the PVOutput forum is on a dumb SMETS1 meter with Octopus and he is getting the same 3Wh metered consumption as you. Powerwall 2 and Smart Meter accuracy (Nov19 post)

Looking at the other posts above there doesn't seem to be any consistency, there's a wide range of steady energy demand from the Powerwalls, between your 3Wh and my 80Wh. I think we have to wait for the experts to find out the cause which having, I hope, stirred up a hornets nest with EDF, will produce results, though quite when is difficult to say.

 

[mikemillar]

I think my meter is SMETS1 (installed several years ago) and I'm seeing at least 0.5 kWh daily drain from the grid even when household demand is satisfied by solar and/or battery. Here's a recent consumption graph from the meter data:
View attachment 521274
I'm not going lose sleep over that consumption as the cost of this electricity is less than half the daily standing charge (25p/day for Octopus Go tariff).

John, Your base metered energy is still in excess of Tesla's specified 500Wh/day. At your tariff the excess may well not be worth worrying about. But from the posts above there is clearly something wrong with some meters and it needs properly investigating. It probably amounts to quirte a lot of money overall. Who knows how many meters are over-reading - it could be many thousands.

 

[Solar1920]

Another example. A day with below average usage and supplied by powerwall all day.

 

[mikemillar]

@Solar1920, so your home consumption on 9 March was 9.3kWh, solar generated 7.6kWh. Therefore the Powerwall should have easily supplied all your home load over the 24 hour period, except maybe for one or two very brief occasions when maybe the demand exceeded 5kW. Even so, your half hourly import never exceeded 0.1kWh. So where does the 1.431kWh usage come from? This is basically my problem and my metered Peak usage is similar to your total usage, my Off-peak usage being recorded separately. So my Peak usage measured by the Smart Meter is quite similar to your total usage.

In addition, I totted up the half hourly usage from your daily usage bar chart and get less than 1kWh, not nearly 1.5kWh. That puzzles me. My assessment is that your meter should not be recording anything like that usage.

 

[Solar1920]

Yes it doesn't make sense to me either, i'd have expected a result similar to Johns of around 0.5kWh.

I've sent the details to my supplier, we'll see what they say!

 

[J_ritchie]

Hi Mike, I was just wondering if you’ve had any headway with EDF? We upgraded our gateway to gateway 2 back in November and switched our SMETS1 for a SMETS2 Landis Gyr E470 Type 5424 with a Octopus for their GO tartif.
We have about 12KW of solar so was a bit surprised to see our meter recording 5-6KW a day at a cost of 80p (including standing charge). I’ve been trying to get to the bottom of it with both Tesla and a Octopus aware of no issues however finding your thread it appears to be quite common.
Just wondered if you’d had any positive or negative responses from anyone? Our old SMETS1 consistently just read zero when we had the original gateway. Thanks

 

[mikemillar]

Well I now have a conclusive result. My friendly EDF manager arranged to fit a check meter in series with the smart meter. This was considerably delayed by the Coronavirus clampdown and was installed a week ago. The check meter is a standard digital Economy 7 meter, also manufactured by Landis+Gyr, who make the smart meter.

After 11 days the check meter ON-Peak reading is still ZERO !! (what I expected since the Powerwall has supplied all the ON-Peak house demand).

The check meter only displays whole kWh's like the smart meter, but I can tell that it has recorded some energy consumption, less than One kWh. The smart meter ON-Peak reading has accumulated 14 kWh over the same period - GAME, SET, but not quite MATCH yet! My EDF manager rang me this morning to say that he is trying to hunt down someone in EDF management who understands what reactive power is, obviously something in rather short supply as I've been raising this possible factor for months!

My assessment is that the smart meter is somehow incorporating reactive power into the ON-Peak readings and may be subtracting it from the OFF-Peak readings. Probably the programmer put a PLUS instead of a MINUS. it wouldn't be the first time in the history of programming. As far as I can see, the total accumulated power readings are quite accurate and agree with the Tesal Powerwall metering figure from the API downloads.

There could be some red faces in high places if this comes out! In fact I don't really think EDF are to blame for this. It's almost certainly the manufacturer, or whoever wrote the firmware progam who is at fault. EDF even seem to farm out the alalysis of their Smart meter readings to a third party, which, I have been told, is why the consumption shown on the website is always 4 days out of date.

 

[Moultymagic]

So .... my story is very similar.

I’ve had solar for about 2 months now and Powerwall 2 for just about a month. Then 1 week after the powerwall was installed I had SMETS2 installed by octopus.

Since the SMETS2 was installed the WiFi display that quotes the grid pull consistently shows 150w to 200w of constant power draw. This happens even when the solar is powering the house or if the powerwall is powering the house (or a mix of the 2). Even if I am exporting to the grid the power to the grid (according to the SMETS2 WiFi display) is 150w to 200w below what the gateway / Tesla app is reporting.

I wrote to Tesla and Octopus and Tesla have put the ball firmly in Octopuses court. A response from octopus says they are working on this known issue with Tesla and that I will not be left out of pocket (I should bloody hope not as it’s over 4kwh per day !!!!!!!)

I really hope this gets resolved ASAP.

I work for a renewable energy company ..... I’m no engineer but I know plenty of the smart fellas that understand reactive power and other power electronics. When I back to work I’ll be asking a few experts the question and get their take on it.

I’ll be watching this space too ....

Good luck all ....