How much does it cost to keep the 2G and 3G networks running? Assume the cost of running the network is borne by the telecoms companies, whereas the smart meter replacements are a problem for the taxpayer.
post
Wasn't 2G switched off a year or two ago?
According to the article, 3G was switched off, while 2G is mandated to be shut down in 2033; but the telcos are preemptively planning to do this in 2029/30.
Nah, I think networks have until 2030-something to turn it off
Consumer phone use will probably get cut off before then, but as the article indicates, there're a lot of non-phone devices that make use of 2G networks too.
I'm probably running on out of date knowledge, but I think I remember reading that some rail signals depend(ed?) on 2G networks, similarly some alarm systems
I think rail uses GSM-R, which is not the same as the general public GSM aka 2G.
Oh really, I thought GSM-R was very much based on 2G, and is/was subject to the same switch off dates
I'm not gonna pretend I know loads about that though
Ah ok some mobile networks have already switched it off for consumers, but maybe not all of them are offline to everyone.
Utility meters last for decades. Making them reliant on a cell network that's constantly changing wasn't such a good idea. They should have licensed some low band UHF frequencies and set up their own network. It doesn't take much bandwidth to read some meters.
It would've made sense if the Meters were made modular with a removable antenna/mobile data modules - rather than having to replace loads of perfectly functioning smart meters because the energy companies wanted to save a few bob 20 years ago.
They did that for the second gen meters.
"The first covers newer, second generation (SMETS2) meters, where only the communications module needs swapping out, but even this is likely to cost £1.3 to £1.4 billion (c $1.7 to $1.9 billion) before the hardware cost of each new hub is even factored in.
The other is for an estimated 7 million older, first-generation (SMETS1) meter installations, which have the cellular comms hard-wired in, so the whole meter typically needs replacing. The report authors estimate this will cost another £2.1 billion to £2.45 billion ($2.8 billion to $3.3 billion), at an indicative replacement cost of £300 to £350 per installation."
That would've the smart choice.
They also could've designed all of this to work as a mesh network so they aren't reliant on anyone elses network.
Yes mesh networks have their flaws but they probably only need to check each meter once a month and it's probably just a MAC address and text. So speed and having the device constantly connected isn't necessary.
It was a rush job, as the old analogue meters depended on a signal piggybacking on BBC R4 LW’s broadcast to switch Economy 7 timings. This ran from the Droitwich transmitter that depended on a specific type of valve which ceased production in the 1980s. In the 2000-10s the BBC bought the last 12 unused valves in the world, but couldn’t make longevity guarantees. The smart meters had to be in place before the transmissions stopped. The transmitter shut down in June.
Wait, why are the individual meters wireless at all?
It seems like if you're running a power cable for ~200 A service (Idk what UK standard is, but that's what I used to in US), running a small fiber cable in parallel would be trivial - and I expect the data rates coming from these would make a fiber connection "future proof" in the most literal sense of the word.
Extreme cost is likely the reason, digging up every power line would get expensive very quickly.
Here in Italy meters talk over the power line itself, like the PowerLine you use in your house. It even allows (but nobody bothers) to connect a user device by plugging it into an outlet (different networks though)
That would mean digging up existing powerlines into every home though. Making them wireless v8a cellular modem makes them independent enough and avoids depending on WiFi which would place the burden of connectivity on the residents.
And it's going to be necessary again in 2040.
However, this is sufficiently far in the future that it is unlikely to trouble policymakers anytime soon.
Lol.
I can't tell if the numbers are bloated for pushing an interest, but with having to spend those billions every 15-20 years, it would probably be cheaper to hire meter readers or pay to keep the 2G available.
My 1950s meter still works fine.
The smart meters were never about reading meters, it's for load shedding. They can turn them off remotely.
But think of the short term profits and dividends for the shareholders!
modular designs prevent this, but nowadays everything must be evil
We're having the same issue in the Netherlands, although the grid managers have started replacing them a while ago. They're being replaced with 4G though, which of course also won't be supported forever.
I think a better solution might be to make a local mesh network on the meters, and one base station nearby that has the internet connection.
Most meters are in areas with plenty of others nearby, and there is already a box thing the grid manager owns somewhere near an electricity connection.
People in remote places could just keep using their existing (or new) meters with 4G.
Then the only reason to replace meters will be because the grid manager themselves wants something changed, for network updates you just replace the base stations.
I would think this is a perfect job for lorawan?
all 12 comments