Smart Meters - Beware!

[mikemillar]

With some reluctance, I have just had a new Smart Meter installed. I did so after assurances that I would retain control over the day/night changeover time and the economy 7 tariff options, and because EDF offers a particularly attractive tariff with a very low off peak rate if one agrees to have a Smart Meter fitted. We have owned a Model S for 4 years and had a Powerwall 2 installed 3 months ago. This is working well and last month we only used 20kWh of peak rate energy at the high EDF rate of 24.69p/kWh. The night rate is 4.9p/kWh on our new tariff, which requires a Smart Meter to be fitted.

The new meter was installed 2 days ago and I immediately detected a problem. With the old meter, the Peak rate digits never budged even 10W from day to day unless the Powerwall ran out of charge. Now I have an apparent constant grid consumption of just under 100W. This is only obvious because I know the consumption should be zero. It doesn't sound much but it amounts to over £150 per year! You might note that 100W over the 17 hour Peak rate period = 50kWh per month - 2.5 times my current consumption! During the 7 hour night period the cost would be fairly insignificant and it would be very difficult to detect as the Model S and the Powerwall charge during this time.

I am currently in discussion with EDF about it. We are going to give it a couple of weeks so that they can get a reasonable period of half hourly readings to analyse. Watch this space!

If it wasn't' for the presence of the Powerwall it's unlikely that I would have noticed anything for some time. There would just have been a slightly suspicious increase in the energy consumption and cost over a period of several months or a year, if one monitors these things, which I do but a lot of people don't. There have been studies which indicate that Smart Meters can over-read by a significant amount, e.g. http://www.yougen.co.uk/blog-entry/2953/Are+smart+meters+overcharging'3F/

This suggests that the electricity companies may be getting an unexpected windfall from the installation of Smart Meters. Of course I'm not suspicious enough to imagine that it's deliberate...

 

[Generator]

Interesting, and easy to,prove one way or another with suitable instrumentation. But, you have to consider the possibility that your new, more recently tested and calibrated meter might be the more accurate of the two. I think the only way I’d get my house under 100W would be to pull the fuses from the cut-out.

 

[mikemillar]

Agreed, it's by no means certain that the old meter, with mechanical drum digits, was accurate at very low consumption levels. I plan to switch off the consumer units as a trial tomorrow and see what happens. I presume I can switch off the Powerwall with the switch on the side, as we did when the meter was installed. And I'll have to turn off the solar as well. The Tesla Model S can also draw a small amont of background processing power when it's plugged in but not charging.

Of course I may have been getting away with 600kWh of unrecorded electricity a year. However, I use PVOutput to log and monitor the Powerwall and that shows basically zero grid use when the Powerwall is supplying the house. I would have thought that the £1700 Tesla Gateway would be at least as accurate as a Smart Meter. The spec quotes an accuracy of plus/minus 0.2%. Resolution on the PVOutput log is 10W, so 100W ought to show up very easily if it's really happening and nothing has changed as far as the Gateway log is concerned.

 

[NewbieT]

Have you tried to identify what is drawing the 100w?

I’m inclined to say the draw might not have been enough to spin the old meter. You can ask for your meter to be sent away and tested but a new meter (tested at the factory, certificate of conformance) probably isn’t wrong and you’ll end up paying a few hundred for the meter test.

 

[mikemillar]

That's certainly possible but as I said, I don't see that a new sophisticated system like the Tesla Gateway would be any less accurate than Smart Meter.

 

[mikemillar]

Also, if the battery is supposed to be supplying ALL the house load, why should there be any power draw from the grid? Certainly not two old light bulb's worth!

 

[Generator]

That's certainly possible but as I said, I don't see that a new sophisticated system like the Tesla Gateway would be any less accurate than Smart Meter.

Sophistication does not necessarily equal accuracy. But meters designed and certified for fiscal use in domestic applications do have a surprisingly broad tolerance requirement. Between +2.5% and -3.5% accurate according to Which, although they don’t specify if that is % of reading or % of full meter capacity. Likewise, the Tesla gateway spec doesn’t say whether the +/- 0.2% accuracy is 0.2% of the full 100A capacity or of the power transmitted at that time, or what calibration standard it can be traced to, if any.

 

[Feathermerchan]

Utilities all have a concern when a customer attaches generation equipment to their grid as you have done. That is should the grid go down for storm, maint, etc your generator can provide enough reverse power to be dangerous to utility workers and other households. So there is always a reverse power relay configured to disconnect your house from the grid. When your system is meeting all the loads in your house, the load thru the meter would be 0 and would be hard to control and keep at exactly 0 as loads are switched on and off. So perhaps the system is designed to keep some small forward power flow to prevent nuisance disconnect from happening.
If I'm all wrong, let me know.

 

[mikemillar]

As I know from research involving measuring very small strains in pressure vessels, pinning down accuracy to NPL standards, as opposed to resolution, can be like chasing your tail!

 

[mikemillar]

Update: Today I turned off everything in controlled manner. I powered the EDF Chameleon Remote monitoring device from a 12-240 volt inverter. Unfortunately, while the BT Hub6 worked on a 12 volt supply I couldn't connect to it (powerline ethernet) so couldn't view my other monitors.

With the solar off and consumer unit turned off and isolated (so no demand from the house) the Chameleon Smart meter power output dropped from 90-100W to 8-11W. Then I turned off the Powerwall and the demand dropped zero. So I assume the Gateway was drawing 11W from the grid - quite acceptable if it can't fully power itself.

Interestingly, during the day, the average grid demand indicated on the Chameleon dropped from about 100W to 60W. The electrician did say that theSmart Meter took a few days to establish the average demand before settling down. Maybe it just can't believe that the house could be drawing zero from the grid during the day! The Gateway may need some form of intelligent neural network based algorithm to decide how much to charge tomorrow, but I can't see why a so called Smart Meter should record other than what is really being consumed. Anyway, maybe i'ts settling down and will record a realistic peak demand in a few days. I can but hope!

 

[Feathermerchan]

AFAIK the 'smart meter' only measures actual consumption/generation. No other funny business.