UK Smart Metering Plan Delayed

Today in Parliament, the UK Secretary for State for Energy and Climate Change – Ed Davey, announced that there would be a one year delay in the GB deployment of smart meters.  That’s not a great surprise to anyone.  Rather than copy what has been deployed in other countries, the UK industry has been developing specifications to make our smart metering system the most complex (and expensive) of any deployment anywhere in the world, and complexity takes time.

However, the timing is not great.  Anyone following smart meter deployment will have noticed a distinct slowdown over the past few years.  Some of the larger US utilities have already rolled out smart meters, helped by $83 billion of stimulus funding.  A significant chunk of Australia has them as well.  But as stimulus funds have dried up, so have the bulk of smart meter deployments.

The one remaining gung-ho project was the UK one, where previous ministers had been keen to pull it forward.  A lot of the rest of the world, particularly other countries within Europe, who are working out how to address the EC directive on smart metering, have been waiting and watching to see how we’re doing.  On the surface it’s looked good.  Underneath there are conflicting interests which have been building up delays.  Unfortunately the timing of today’s announcement plays to quite a lot of those underlying politics, which could further derail the future of the program.

It’s never been clear to me why the UK needs such a complex deployment.  The fundamentals are sound – it helps to produce accurate bills, provides insight to help utilities manage their business (if they ever work out how to use it), and the decision to provide consumers with In-Home Displays should help educate them and reduce consumption.  But that could all have been delivered at lower cost with the type of meters and energy monitors which were available six years ago.  Instead, the UK has developed a complex variant of the ZigBee 1.0 Smart Energy Profile to add prepay, support for vastly complex tariffs and historic data storage, which has taken years of development and is still very much a work in progress.  All of which is making our meters much more expensive, and pushing their delivery dates out into the future.

Behind the meters, DECC has constructed a body responsible for connecting all of these meters, collecting readings and then passing them on to the relevant owner.  It’s called the DCC – the Data and Communications  Company.  Once formed, this will award contracts to CSPs – Communication Service Providers, to implement the actual connections to the meters, as well as providing interchangeable comms hubs for each meter.  The argument for all of this is that it allows consumers to switch supplier easily.  It’s a massive IT project of the kind that Governments love, because it make it look as if they can specify complex IT, and which suppliers equally love, because they know that Governments will get the specification wrong, leading to blank cheques being written before the inevitable cancellation at some point in the distant future.

No other country has felt the need to do this, but then Britain didn’t get to be where it is today without a long history of screwing up major IT contracts.  Because of the complexity of the DCC and CSP operations, we’ve been through interminable rounds of negotiation as bidding suppliers try to decrease the requirements laid upon them, allowing them to make a bigger profit, along with a “get out of jail free” card for when it inevitably goes pear-shaped.

That’s a large part of the reason for the delay.  But it’s not the only one.  The complexity of the meter specification has ballooned, with those involved playing the old specmanship game of seeing who can piss highest up the wall.  In an industry which is not used to wireless protocols or interoperability, they have yet to learn that simple is best.  Instead they’ve concentrated on turning meters into complex computing devices rather than just connected meters.

You can put part of that down to an industry that is suffering the pains of having technology thrust upon it.  There is no doubt that much of it has been mugged by technology providers and consultants.  But it’s probably also a reaction to low cost smart meters coming from China.  A complex, moveable specification acts as an effective trade barrier, which is why industry players will spend time and money developing it.  But it also contributes to today’s slippage.

These delays and changes don’t play to the business model of utilities.  They’re very risk averse and one thing they hate is spending money which might be wasted.  One of the worst wastes in their eyes is that of stranded assets – deploying a meter or another piece of equipment which needs to be changed or replaced within its working life.  When they put a meter on your wall, they’d prefer to leave it there for 25 years or more.  (My gas meter, which was replaced a few years ago, had been on the wall for 59 years.)  They’re scared stiff that these new-fangled meters might not last fifteen years.  Even worse, an early deployment of smart meters might need replacing in five.  Faced with that concern, they’d rather not deploy anything.

The purpose of the initial Foundation stage of the UK deployment was to let utilities start deploying meters so that they could learn from that experience.  Learn about the issues of installing connected devices, about commissioning the wireless comms, about remotely managing them and about coping with the volume of data pouring into the backend, without having the complexity of the DCC being in the equation for this phase.   But with a spec in flux and a constant desire to add complexity to the system, all that most of them could see was the spectre of stranded assets.  So there has been an underlying pressure to delay.  That has fed a vicious circle of equipment vendors trying to outspec each other in the hope of winning contracts.

Part of the Government announcement today was an attempt to reassure the UK utilities that they need not worry.  There’s a clause in the document that would allow the Foundation meters, which confirm to the earlier SMETS1 specification, to be retrospectively connected to the DCC.  However, it places this responsibility of providing a future backwards compatibility on the DCC and it’s not clear that anyone trusts them to deliver that level of interoperability.  So the statement today, whilst appearing well-meaning, may have the opposite effect.

What the Government does not understand is the potential consequences of having let this happen at this stage.  There are two risks, both emanating from the larger political games being played within elements of the industry.

The first is a local one.  The GB metering specification uses ZigBee Smart Energy Profile (SEP) version 1.x for the smart metering network within the home.  It currently operates at 2.4 GHz, as it does everywhere else in the world.  However, in the UK we have a lot of buildings with thick walls, and a fair proportion of high rise flats, neither of which allow 2.4G Hz to propagate from meter to in-home display.  That’s not a great issue in other countries, as they don’t give every consumer a free in-home displays.  Because we do in the UK, and had to include the energy savings they’ll allegedly deliver to justify the cost benefit of smart meters in the first place, it means that around 30% of homes need another wireless solution if their in-home display is going to work.  Otherwise the economic justification for the deployment falls over.

The current approach to that is to develop a new version of the ZigBee standard which operates at 868MHz.  It’s technically possible, but writing wireless standards takes time.  There was no hope that it would be ready for the original mass rollout in 2014.  Nor will it be ready for the new date of 2015.  But it might just be in 2016.  If utilities start their roll-out in 2015, as currently planned, then they will need to start with 2.4 GHz meters.  They’ll then need to move to dual-band 2.4 GHz / 868 MHz meters and comms hubs from 2016 to provide backwards interoperability.  Which, for their business model, argues that they should delay any roll-outs until 2016 so that they only need to supply 868 MHz equipment.  The problem is that if the new standard is delayed, as it inevitably will be, nothing will get deployed until 2017 or later. However, as long as they believe their vendor’s PR, which says 868 MHz development will be trivially easy, then any delay to the program now makes that option of waiting for 868 MHz look very attractive to a utility.

The other problem is that most of the rest of the world is moving to a different version of the ZigBee Smart Energy Profile, called SEP2.0, which runs over IP.   The UK Government chose not to adopt it, because they initially planned to start deploying meters in 2012, which was well in advance of when SEP2.0 would be complete.  However, last week the ZigBee Alliance announced that the SEP2.0 standard had been released.  It means that meter vendors around the world will be turning their development teams towards products that use SEP2.0.  The longer that the UK delays, the greater the pressure will be for it to migrate its specification to SEP2.0.  As SEP2.0 is incompatible with SEP1.x, any SEP1.x products deployed ahead of that change would be stranded assets.  Which again suggests to utilities that they should sit and wait.  And it’s an industry which is exceedingly good at doing that.

And while the UK prevaricates, so will the rest of the world.  The UK deployment was probably the most high profile deployment, not least because it was going to demonstrate that complexity could be managed.  The larger vision of the program also resulted in the formation of a number of innovative start-ups developing in-home displays and home energy management products.  They are now at risk of going bust if there is a twelve month hiatus in orders.  Even larger players will start to look at the value of continuing investment to support the program as they see any revenue potential disappearing into the distance.

It’s not the death of smart metering, but it will be a shock that will reverberate far beyond the UK.  If the UK is going to claw back its reputation as a pioneer, it may need to radically reassess how it is going to deliver smart metering.  The smart metering and consumer engagement landscape is changing.  Any delay may mean that existing meter and engagement designs may look obsolete by the time that they come to be deployed.  So rather than being a breathing space, this may be better interpreted as a call for Government and industry to reflect on their fundamental ideas.