New Report Examines the UK’s Public Switched Telephone Network Switch Off
A new independent research report has been published by BroadbandSwitch.uk, which provides a fairly detailed and useful overview of Britain’s move to switch-off the old analogue phone network on 31st January 2027, which was previously delayed from the original end of 2025 deadline in order to protect vulnerable consumers.
Just to recap. Openreach are withdrawing their old Wholesale Line Rental (WLR) products as part of this change, while BT are retiring their related Public Switched Telephone Network (PSTN). But at present around 1.9 million UK lines are yet to switch to a digital alternative and the focus of the remaining effort is largely on helping those who use vital home telecare systems (e.g. elderly, disabled – vulnerable users), which aren’t always compatible with newer IP-based phone services (telecare providers were slow to adapt).
The Last Dial Tone paper doesn’t say anything that we haven’t reported on before, but we do think it’s one of the better overarching summaries. The report’s modelling also estimates that a residual tail of between 150,000 and 400,000 lines will not be migrated in time, although Openreach are expecting something more in the high tens of thousands.
Advertisement
The good news, as recently reported (here), is that consumer lines won’t face a hard disconnection on the deadline and there will be an emergency solution for those who fail to migrate. Business lines, however, won’t benefit from the same protections and will need to get a move on.
Dr Alex J. Martin-Smith, Report Author, said:
My hypothesis, stated plainly so it can be marked against events: the headline date will hold, but completion will be achieved partly on paper. A residual tail of lines, our modelling suggests somewhere between 150,000 and 450,000, will not be migrated in time. Instead they will be caught by transitional products that BT and Openreach have quietly pre-built for exactly this purpose: an emergency-calls-only line called EVAc, and analogue-style stopgaps known as PDPL and SOTAP that run to around 2030. The platform retires on schedule. A minority keeps a dial tone through the side door. That is a soft landing, not a clean finish, and not a delay.
Why lean this way? Three reasons run through this report. First, the network is failing in plain sight: it consumes around 1% of British energy usage, suffered more than 2,600 major incidents in a single year, and runs on parts nobody manufactures any more. Second, the date is a commercial decision, not a law, and the government has said it will not legislate a pause. Third, and decisively, the escape hatch already exists. The argument that forced the 2023 pause, that vulnerable people would be cut off, has been structurally defused by products designed to make sure nobody is.
We put the probability of this soft landing at 72%, a formal further delay at 23%, and a genuinely clean completion at just 5%. Part Three shows the working, and page 26 publishes the exact criteria by which we will score ourselves after the event. Authority is cheap to claim and expensive to demonstrate. We have chosen the expensive route.
Mark is a professional technology writer, IT consultant and computer engineer from Dorset (England), he also founded ISPreview in 1999 and enjoys analysing the latest telecoms and broadband developments. Find me on X (Twitter), Mastodon, Facebook, BlueSky, Threads.net and Linkedin.
« 140 Government Funded 4G Mobile Mast Upgrades Now Live Across UK
Advertisement
Leave a Reply
Privacy Notice: Please note that news comments are anonymous, which means that we do NOT require you to enter any real personal details to post a message and display names can be almost anything you like (provided they do not contain offensive language or impersonate a real person's legal name). By clicking to submit a post you agree to storing your entries for comment content, display name, IP and email in our database, for as long as the post remains live.
Only the submitted name and comment will be displayed in public, while the rest will be kept private (we will never share this outside of ISPreview, regardless of whether the data is real or fake). This comment system uses submitted IP, email and website address data to spot abuse and spammers. All data is transferred via an encrypted (https secure) session.
The area of remaining concern is emergency phones in lifts, especially in small businesses and sometimes also domestic where these are easy to forget, and similar use cases.
I would hope that existing lift safety inspections do check the alarm phone works, as a backstop.
I would also hope that there is a standard universal 4G module that can be retrofitted quickly and cheaply (instead of changing the anaalogue line to VoIP with battery backup).
The report does cover this but I suspect underestimates (12k) the population of life lines remaining.
The reference for this figure in the report is an ISPReview post from Frbruary which in turn seems to be quoting Openreach (and I haven’t managed to find the true source of the 12k estimate)
Anybody operating a lift without a service contract and monitoring is acting negligently, anybody taking money for a lift monitoring contract that is going to have issues with the phone line being withdrawn is borderline committing fraud, as they’re not providing the service they’re charging for.
Sometimes I feel that BT would have attracted less grief if they’d set a 10 year date for PSTN closure and not tried to offer a replacement product. The majority of the friction seems to be in the switch to Digital Voice, but if the messaging was “your care alarm will not work in 2027” then it would have been solved by the alarm providers replacing the base unit for a Wi-Fi/cellular connected unit.
@jonny you are of course legally correct, but we know that
a) businesses do not always discharge their safety responsibilities properly or trade honestly (and domestic lifts also exist)
b) there is a lag between annualized inspections, and a lift line could be out of service for several months due to analogue connection being withdrawn before it is detected by an inspection. Unfortunately lifts don’t “fail safe” for lack of a dial tone. So there is a risk that could become a incident, penalized after the fact, and that’s not great.
“…4G module that can be retrofitted quickly and cheaply (instead of changing the analogue line to VoIP with battery backup…”
Wouldn’t the 4G module need a battery? Loss of power at the lift was mitigated power over PSTN. A good mobile phone battery will last a couple of days. Maybe what we need is battery backups for ONTs / routers that only power up when actually being used, rather than only managing to keep it alive for a couple of hours. The same would benefit care alarms of emergency calls for households. Allow the battery full power only to be triggered if a live signal is sent from the router and the router only use its battery if the connected handset is operated (kill the WiFi and main internet access during emergency mode)
Modern monitoring should probably be continuous, and that connection being dropped should flag as an issue, as should power failure (with a small battery to ensure this can be sent and allow for cons with anyone trapped in the lift
We need to move on from the idea that it all has to be make a call when an event has happened. That made sense 30 years ago, but IMHO not now with packet switched networks
This is exactly the kind of discussion I hoped the report would prompt, so thank you all for it. The lift question is the right one to be worrying about, and you’ve each put your finger on a different part of it!
MilesT, you’ve traced that 12,000 to the same place I did, and I’ll be honest: the trail goes politely cold. It’s Openreach’s own February 2026 statement, reported through ISPreview, for lift lines not yet migrated at that point, and I haven’t found an independent count behind it either. so I’d treat it as a floor rather than a firm total, and you’re very likely right that the true number is larger. Pinning down a better-sourced figure is exactly what the gap register and the corrections log are for, and I’ll add it to the list. Thank you for pushing on it!
On the safety risk itself, I think you’re completely right, and it’s the part that worries me most. Lifts don’t fail safe for the want of a dial tone, and the lag between annual inspections is exactly where a quietly dead line becomes an entrapment nobody saw coming! It’s why lifts sit in the critical band of the report’s severity map, and why EN 81-28 requires two-way emergency voice that keeps working in a power cut. Jonny is right in principle that monitoring should catch it, but MilesT’s point about real-world lag, and domestic lifts that nobody is contractually watching, is the practical reality.
On the retrofit and the battery, Just a thought has it spot on. A 4G module is not a free pass, because it still needs its own power. The genuinely clever thing about the old network was that it was powered down the line, so resilience came built in, and every digital replacement has to rebuild that deliberately. The 2022 update to EN 81-28 actually makes the point for you: it now requires battery back-up on the cellular gateway itself, not just the alarm unit. And your idea of a battery that draws full power only when a call is live, rather than trickling away, is a genuinely good one. The one-hour minimum that providers must offer at-risk households is widely thought too short, with MPs arguing for four to six, so smarter power design is exactly where the sector should be looking.
This is the conversation that actually matter (it really does!), far more than the headline date. If any of you spot something in the report that’s wrong or could be sourced better, the corrections log is there for exactly that, and I’d be glad of it – please do message me [email protected]. Thank you again.
When I queried if there was a “drop in” 4G replacement (as an “industrial” module), I had presumed (but did not state explicitly) that the module would include as standard a multi-hour battery backup, continuously charged from the power supply for the lift itself.
This would put the connectivity in the same profile as to expectation for VoIP landlines, especially domestic, where the cutover expectation is “people will use mobile phones if the landline is unavailable, and is also similar to the 4G enabled security pendants that have become commonplace in the last few years (with a multi-network SIM card).
Potentially the security pendent type of device can be integrated to the lift (ISTR having seen a device that has connectors for external hardwire call button and two way microphone and speaker), in a dedicated (to the lift) or shared mode (also used for pendant, in a domestic situation)
I wonder will they install PDPL into smaller exchanges. There are many in the Highlands with a few hundred customers, maybe some will only have one or two holdouts
Good question, and a sharp one. One wrinkle worth knowing first: PDPL is actually closing to new sign-ups around mid-2026, so the real catch for late holdouts is EVAc, the residual emergency line that exists precisely so nobody, however remote, is simply cut off. Openreach’s own phrase for it is “no big bang.”
That said, you’ve found a genuine edge case. EVAc moves lines through a gating process rather than automatically, and capacity applies, so how it behaves in a Highland exchange with one or two stubborn holdouts is exactly the kind of operational detail nobody has published (i think?). it’s on my list to ask about. Thank you, that’s a good one!
afaics the ‘1% of British energy usage’ is 0.8 and covers the whole BT group. Is there a number that actually covers what they think the old -v- future system consumes?
I’ve been wondering about this too – it’s also the figure that was mentioned in the Zen interview with someone from Openreach published a few days ago.
A bit of searching led me to the report linked below, which says “The electricity
used to maintain the UK’s PSTN is estimated to be 313 GWh per year across the UK’s PSTN
providers (KCom, Virgin Media O2 and BT Group). This accounts for over 0.1% of the entire UK electricity consumption each year.” (section 4.3.1 on page 20). I think the Openreach person from the Zen interview probably meant to say 0.1%, although to be fair to them they did have a lot of facts and figures, and we are all only human!
https://www.bt.com/bt-plc/assets/documents/about-bt/all-ip/migrating-from-analogue-to-digital-landlines.pdf
Also reading a bit further in that report, section 4.3.2 on page 21 covers power usage by FTTC/FTTP: “According to BT Group, FTTC broadband consumes 2.4 watts per customer, whereas full fibre consumes just 0.3 watts per customer to provide a much higher capacity service.”
Don’t forget that 0.3 watts doesn’t include power consumption of the ONT which I understand is somewhere around 7 watts. To a certain extent they’ve simply shifted the cost of running the network from themselves onto the customer.
This is exactly why I put the numbers out in the open, and threads like this are the best part of it.
You’re all right, and Al got us started. The BT report you linked has the better figure: the PSTN draws around 313 GWh a year, just over a tenth of 1%t of UK electricity, and that’s the one i’ll carry from here. It’s a lovely example of how this is meant to work, a single line from an interview gets held up against a primary source, and between us we land somewhere closer to the truth. Into the corrections log it goes this week, credited to this thread, and my rather over-excited iPhone line gently deflates alongside it.
Dave, yours is the point I keep turning over. That 0.3 watts is the network’s figure and quietly leaves out the ONT humming away in your hall,so some of the running cost has simply moved from BT onto the rest of us. That’s a genuinely interesting shift, and exactly the sort of thing worth pulling on further.
So please, keep it coming. Half the value of a report like this is the conversation it starts, and I would far rather share everything in the open and be sharpened by people who clearly know their subject. Thank you, all of you. This is the good bit.
People may want to exercise caution whilst reading this report, I am not convinced by its trustworthiness. Of course, this is just my opinion, but I think it would be fair to explain my reasons for this are:
– Whilst the report appears to be well sourced, some of the data cited seems to be irrelevant and included to try to support a point that the data isn’t relevant to. Perhaps to just pad out the report? For example, it twice mentions fines for some providers related to 999 failings (which are unrelated to the PSTN migration) in the same context as VMO2’s fine for its PSTN migration failure that sadly led to two deaths. The report tries to link 999 failures to VMO2’s telecare failure when these happened for very different reasons and in very different contexts. In the second mention of these fines, the report only states in very small light grey disclaimer that only one out of the four listed fines was related to the PSTN migration.
– The report makes a big deal of trying to say why it’s trustworthy, but at least some of the author’s degrees are all from a diploma mill (“LIGS University”, which is not a properly accredited educational institution).
– The writing has serious AI slop vibes in a lot of places, again not exactly a sign of trustworthiness.
– It uses an awful lot of words to say what could have been said concisely in 5-6 pages (and repeats itself), which to me suggests quantity over quality.
Indeed the whole thing absolutely reeks of AI waffle. Moreover, while you can reasonably model how many lines will be unmigrated by the cut-off date the attempt to derive such precise probabilities for future human decisions is not credible.
Thank you for reading it, and for taking the time to comment. I mean that.
You’re right that I didn’t study at a famous university. Not everyone gets that chance. What I’ve done instead is leave a good career, work hard at this, and try to let the work earn its trust rather than the letters after my name. Every figure in the report is dated and sourced so anyone can check it, and we keep a public corrections log precisely so people can tell us when we’ve got something wrong and we can put it right. I’m sure it isn’t perfect, and I’d genuinely welcome being shown where.
But honestly, none of this is about me. It’s about the people the switch-off affects. The family wondering whether their phone will still work in a power cut. The small business or charity with an older system and no fallback. And the 1.8 million people who rely on a telecare alarm, for whom this is a safety matter, not a tech story. That is who the report is written for, in plain English, on purpose. I would rather over-explain and be useful to someone meeting this for the first time than be brief and leave them behind.
And there is a good story underneath it all. Britain has quietly built one of the best full-fibre networks in Europe, faster than almost any comparable country. The old copper line is being retired because something genuinely better has taken its place. The way I see it, my job is simply to help people and businesses make the most of that, fairly and without the jargon.
So I’m not really interested in scoring points. I care about how we move forward and get this right for the people who depend on it. If you’d like to talk through any of the detail, I’d honestly value it. I’m easy to find: linkedin.com/in/alexmartinsmith.
@Alex Maybe don’t use AI to write comments for you? I’m not getting into a discussion where the replies are just some copy/pasted AI slop.
@Anonymous
No worries, I won’t drag you into a back and forth you’d rather not have. I’m honestly not claiming to be the expert here, I just care a lot about getting this right for the people it affects, and yes, I do use AI as a tool sometimes, which we say plainly on the site, and I find it genuinely useful as long as the thinking and the responsibility stay mine. On the fines, you’ve got a fair point: only one of the four really belongs with the migration itself. I’d grouped them because the switch-off carries a 999 dimension of its own, but you’re right that the distinction is worth drawing cleanly, so I’m separating them out in the report and updating our corrections log to reflect it. Credit to you for flagging it. The offer to talk stands if you ever fancy it, no pressure at all: 0330 122 1223, [email protected], or linkedin.com/in/alexmartinsmith. Thanks for taking the time to read it, and I genuinely wish you well Anonymous
There is a risk they cause increasing disruption in the haste to hit this deadline. My mother in law is currently on Plusnet Broadband and phone and is being moved to EE. They just sent a letter and informed her they were sending a new router in the post. She’s 85 and would have no idea what to do with it especially as she has no power socket near her phone (my father in law did a Heath Robinson extension for the router!). Her mobile signal is woeful on all networks so she depends on WiFi calling at home. I am having to call them up to send an engineer and I’m insisting on a battery backup as she has no mobile signal.
Darren, it sounds like you’re already doing right by her, and pushing for the engineer and the battery backup is bang on.
She’s on firmer ground than that letter makes out, mind. The battery backup isn’t a goodwill gesture, it’s something they have to give her, free of charge. Ofcom’s rules say anyone who relies on their landline and can’t fall back on a mobile has to be given a way to still reach 999 in a power cut, and that’s exactly her situation. So don’t let them talk you out of it.
The engineer is the same story. Posting a router to an 85-year-old and leaving her to it really isn’t good enough, and as a vulnerable customer she can have someone come out and set the whole lot up. While you’ve got them on the phone, get her added to their priority register too, so that extra help is on record next time round.
The bit I’d really press EE on is the wifi calling, because it’s what trips people up. If the power goes, her broadband goes with it, so she’d lose the landline and the wifi calling in the same instant. For her, then, that backup isn’t a nice extra, it’s the one thing keeping a working phone in the house when the lights go. I’d ask them flat out what it actually keeps running, the phone or the whole router, and how long for. An hour is the minimum they have to offer, but with no mobile to lean on, I’d be holding out for longer.
For the ‘no socket near the phone’ business, an engineer can usually sort it with a cordless handset sat next to the router, so she keeps a phone where she actually wants one rather than where the wiring happens to be. Might let your father-in-law’s extension off the hook while you’re at it.
And you’re right to be wary of the rush. These protections are there for people exactly like her, but they’re only worth anything if someone makes them happen, which is what you’re doing. Good on you.
Recently had a power cut – first since our phone line forced to switch to IP based. Router not connected to a UPS so unable to make landline calls. No problem I thought, I will use my mobile … Discovered no mobile signal, appears the phone mast I connect to has no battery backup. Looks like I’m getting a UPS as a precaution.
Should your power provider not supply a BBU for when they fail to operate a robust and resilient service? Every time I read that the telco should act as the provider of the power back up solution, this question always comes to mind.
My thoughts also. A lot of people go on about power cuts as if they’re a daily occurrence. Either that’s a massive exaggeration or they need to go to Ofgem (not Ofcom) and get some answers as to why that is the case, and then get a generator.
Often its people claiming that rural mains power is less reliable. I grew up in a rural area in Cornwall and I’d often need to replace the UPS’s batteries before they got a proper workout from a power cut.
The first long-term power cut I experienced, aside from it being off during the day for planned maintenance, was in the January storm. Discussion on BBC local radio turned to landline phone reliability and of course neither the presenter nor the callers (hmm) were in possession of the facts.
Exactly, that’s why my chosen commute is a Challenger Mk 3 . . just in case you understand.
Don’t upset Ms Kirby . . she’s got enough to do commuting between Windsor and Sweden and squirrelling away those outrageous emoluments, . . . without engaging with real people’s telecoms problems. Protect the Queen Bee at all costs went the exhortation from middle management otherwise we’re all in the surcharging stuff.
@Dave, that’s the whole thing in one story really, and I’m sorry you found out the hard way. The part people miss is just what you hit, it isn’t only the router at home, it’s that the mast itself can be every bit as dead as your house. Ofcom looked at this last year and found only about a fifth of mobile sites have even an hour of battery behind them, and only around one in twenty can manage six. So “I’ll just use my mobile” quietly stopped being a safe bet the day the line power went. A UPS is a sensible move, for what its worth.
Ex Telco, that’s a fair question, and one I’ve chewed on myself. The reason it lands on the telco rather than the electricity company is that the old copper line carried its own power down the wire from the exchange, so it kept working when your house didn’t. That’s the bit that’s been taken away. Ofcom’s logic is basically, if you’ve removed the thing that used to keep people connected in a cut, then you’re the one who has to put something back. The reliability of the mains itself though, you’re right, that’s not really the telco’s to answer for.
Ivor, I’m with you that power cuts aren’t a daily event for most of us, and Ofgem is absolutely the right door to knock on about why the mains drops. I’d just gently say the worry was never really about how often. Its the one long cut, the storm you mentioned, when the mains and the masts tend to go down together, and the person who used to have a phone that always worked suddenly doesn’t. A generator or a UPS sorts it for those of us who can rig one up. It’s the ones who can’t, the 85-year-old on her own, that the whole fuss is actually about.
Properly good thread this, the kind I learn from. Keep it coming!
Power cut experience, North Norfolk (town rather than truely rural, wityh some overhead cabing). At least one per quarter.
Most only a few seconds (enough to require resetting the clock on the oven which is irritating),
Maybe once a years a cut of a few minutes (varying seasons), but long enough to be material if unlucky in the timing in needing to make an emergency call.
With ONTs/Routers/VoIP (and no battery backup), the “few seconds” would extend to a few minutes while everying reboots and reconnects, i.e. more opportunities to be long enough to be material in the timing in needing to make an emergency call.
According to a Poster at Reading Station
there will be a Hard Cut Off
on the 27th of Jan 2027
If one didn’t know better, I’d say its all going according to Uncle Klaus’s plan, . . cast everybody except the extremely financially well enabled off the lifeboat . .accommodation,health,schooling, finance and banking and telecoms.
Where’s the energy saving . . .’Service’ users will be swapping landlines for mobiles . . and its reported that some of the mobile chargers are left-on 24/7 . . . it could shift the AMOC veer point another 25 miles north.
“Nothing to see here” . .move along as the overly sanguine marketing boys just down from Camden and Somerset are oft to say . . . unless you personally get the shitty end of the stick.
The madness that is BT Group. The in-laws have Plusnet for broadband and an analogue line. Plusnet (part of BT) doesn’t do Voip so they are forcing them to move to EE and a digital voice line. OK, only slightly silly because why doesn’t BT give Plusnet a Digital Voice service to add to their product set? No idea but that’s how daft BT are.
However, it gets worse. Having been told they have to move and been given a date for an engineer to come and set it up (they are late 80s and one is going blind) they then have a call from Plusnet saying “do you want to keep your analogue line? So what are going to do? Charge them for an analogue line for the next 7 months, then cut it off? Utter Utter crazyness. Probably mis-selling too.