Saturday, April 8, 2017

Dumb meters and dumber journalism

This post was inspired by a bunch of press coverage last month of a Dutch research paper, which found big differences between the measurement of certain "smart" meters and a regular (old-style) electromechanical meter. Here is one of the more sober reports. Choice quotes:

Smart meters are giving readings up to six times higher than the energy consumed by households when connected to energy-saving light bulbs, according to scientists.[...]

It is the first ever proof that smart meters, which the Government wants in every household by 2020 to improve the accuracy of people's energy bills, are producing readings which are too high.[...]

So called "green" devices such as energy saving light bulbs, heaters, LED bulbs and dimmers change the shape of electric currents which can result in a distorted reading, it said.

The whole article illustrated by a picture of a pensioner's hands, clutching on to the last of their meagre savings in front of a log fire. Heartrending stuff, obviously. But nothing on the really hysterical coverage:

Smart meters can give readings almost seven times higher than the actual electricity consumed – particularly in homes when energy-saving bulbs are used, a study found.

Modern devices including dimmer switches and LED bulbs can confuse some smart meters, leading to massively inflated readings and higher bills.

All this is based on a paper published in the IEEE Electromagnetic Compatibility Magazine, from research conducted in the Netherlands. The University of Twente's press release is here, and the abstract is here. (In theory you can buy a PDF of the full paper for the modest donation of US$33, but the IEEE's portal defeated my best efforts to give them money. I eventually managed to read the paper from my local academic library.)

Now, the first thing to note about the above coverage is the heavy emphasis on household bills. The Telegraph article mentions the word "household" twice in the first three paragraphs, and backs it up with that scary picture. The Mail unabashedly talks about "homes" and "massively inflated readings and bills". But as even the paper abstract makes clear, it's not about household meters. Your average household has a single-phase electricity supply and a single-phase meter to go with it. Here is what the paper has to say about single-phase meters:

Several single-phase static energy meters were measured in various setups. [...] The results can be summarized in one sentence: no deviation beyond the specification could be observed; no influence of interference due to interfering or distorted voltage, and no influence caused by interfering currents were observed.

No, all the headline-grabbing results concern three-phase meters, generally used in large commercial and industrial premises. That poor old pensioner with the fire? Not affected. "Homes where energy-saving bulbs are used"? Not affected. (Well, to be completely accurate, it's not unheard-of for a home to have a three-phase supply. But unless you've got a welding station in your garage or something, it's pretty unlikely.) The household meters? - those took everything the researchers could throw at them, and ticked along like clockwork.

So, about those three-phase results. The Torygraph makes the connection to "So-called "green" devices such as energy saving light bulbs, heaters, LED bulbs and dimmers". In fact, to get the big anomalies, you need to be using both a dimmer switch, and energy saving (CFL/LED) bulbs - and not just a handful of them either, you need dozens of the things, all connected to the same dimmer switch. And then you need to turn the dimmer switch to 135°. If you leave the dimmer switch alone, or turn it to a mere 90°, you actually save money. Here is the, pardon the pun, money chart from the paper:

DimmerMeterResistiveCFLCFL+LEDCFL+LED+Resistive
SM1-2%-4%-4%-3%
SM2-3%-9%-11%-3%
SM3-3%-7%-6%-3%
SM4-3%-7%-6%-4%
45°SM1-14%0%-4%-16%
SM2-14%6%-5%-16%
SM3-3%7%-8%-3%
SM4-4%7%-6%-3%
90°SM15%-46%-52%-40%
SM25%-46%-53%-40%
SM3-1%3%1%-6%
SM4-2%-26%-28%-20%
135°SM1122%253%169%265%
SM2105%268%180%276%
SM3-1%-10%-3%-4%
SM42%-46%-36%-39%

So you need a three-phase meter and a whole lot of low-energy bulbs that are connected, in series, to a single dimmer switch that for some reason is permanently melted into position at 135°. And then you'll get inflated readings?

Well, maybe. It's still far from a sure thing, because only some three-phase meters show this effect - and the researchers, bizarrely, decline to name them. They do say that the offending meters use Rogowski coils (a century-old technology) for measurement. But no-one seems to know, and the researchers aren't telling, which meters those are.

Which brings me neatly to my topic for part 3 of this series: what really is wrong with smart meters. For now, what I can tell you for sure is that if your bill has gone up since you got a smart meter installed? - there is no reputable evidence to support the idea that it's the meter's fault.