3 Roll-out stages and timescale
37. Roll-out will happen in two stages—foundation
stage and mass roll-out, when the vast majority of smart meters
will be installed. Foundation stage, which is currently under
way, is seen as a period for learning about smart meter usage
and dealing with teething problems. Energy suppliers have taken
very different approaches to installing smart meters during this
stage. When we asked how many they had installed to date, the
figures they gave ranged from 5,000 to 1 million.[81]
Mass roll-out is set to take place between autumn 2015 and the
end of 2020, but when this inquiry was launched, mass roll-out
was due to start at the end of 2014 and finish by the end of 2019.
On 10 May 2013, DECC put the dates back by a year in response
to concerns about timescale.[82]
Concerns about the timescale
for mass roll-out
38. A number of witnesses warned of the potential
consequences of pressing ahead with mass roll-out, particularly
with the 2014-2019 dates, before certain requirements had been
met.[83] National Grid
outlined its concerns as follows:
Feedback gathered through our pricing consultation
regarding the smart metering mass roll-out has indicated a general
view that completion by the end of 2019 remains highly challenging...The
uncertainties that still exist regarding technical and infrastructure
requirements result in a slower roll-out profile than currently
expected…with smart meter roll-out completion taking several
years longer than currently forecast.[84]
39. Others raised concerns that if important technical
and infrastructure requirements were not in place before roll-out,
costs could increase significantly and some consumers could have
a poor experience, which might have a reputational impact on the
roll-out programme.[85]
EDF and SSE described the potential for a "delivery bubble"
towards the end of roll-out, with the cost of installing meters
increasing as the deadline for completion approached.[86]
EDF outlined the risks and the choices facing suppliers in this
way:
As the 2019 completion date is a Licence Condition,
Suppliers have two choices: either to start the rollout ahead
of the delivery of the key enablers, or delay rollout and compress
the delivery period. Both choices are likely to increase the costs
to consumers, deliver a sub-optimal customer experience and introduce
unnecessary risk to the GB programme and the delivery of the expected
benefits.[87]
40. Andrew Ward of ScottishPower gave an example
of how consumers could be affected if problems were found in smart
meters after they had been rolled out on a wide scale and these
problems then had to be rectified by reinstalling meters:
Of the initial 30,000 meters that we deployed
in 2010, we have had to replace 5,000 of the SIM cards that are
in those meters. The understanding we had when we installed the
meters was that the SIM cards would be sufficient to last the
life of the meter, so that has gone wrong. We have had to interrupt
the lives of 5,000 customers and reinstall those meters.[88]
Witnesses were particularly concerned about:
- the fact that the technical
specifications that will ultimately need to be met by smart meters—SMETS
2—have not yet been finalised;[89]
- the need for the Data Communications Company
(DCC) and communications system to be up and running for SMETS
2 meters to be fully operational;[90]
and
- the need for proper end-to-end system testing
before mass roll-out.[91]
We consider two of these key issues—SMETS 2
specifications and the DCC—in later chapters. The need for
proper system testing is discussed below.
System testing and analysis
41. Many witnesses highlighted the importance of
ensuring that the smart metering system and smart grid were secure
and that sufficient time was allowed for end-to-end system testing
and analysis.[92] The
Institution of Engineering and Technology (IET) said:
Throughout the programme, the IET has repeatedly
stressed that secure operation of individual components of the
smart metering system, though important, cannot guarantee system
security. End to end system security is critical...The tight time
constraints should not be allowed to compromise rigorous end to
end security analysis and testing of the resulting system.[93]
Dr Martyn Thomas of the IET emphasised the need for
rigorous analysis "using mathematically formal specifications
and the associated tools" to establish whether there were
"vulnerabilities that could be exploited, or combinations
of circumstances that might cause a significant failure, which
would only appear later on and which would then cause a need for
substantial rework."[94]
He also noted that such analysis was "not particularly expensive",
had "reduced the final cost of systems" everywhere it
had been used, and would "reduce the testing times".[95]
42. Andrew Ward gave a practical example of what
could go wrong:
The risks are real...I will give you an example
from part of our global group in America. They have now installed
over 600,000 meters and I believe the common misconception is
that software upgrades on the meters can be done electronically
from a distance—you don't need to attend the property. As
part of that deployment they rolled out, at the point of 200,000
meters they had to replace 5,000 meters because they could not
update the communications over the wire. They had to again attend
that property, a physical visit. It is a real example of what
could potentially happen in the UK. That is why there is a desire
from a ScottishPower point of view to test thoroughly what is
actually in there before we mass deploy in the UK.[96]
NEW MASS ROLL-OUT DATES
43. Several witnesses welcomed the new 2015-2020
timescale for mass roll-out.[97]
Tony House of SSE said it would enable a lot of the risks that
had been identified "to be mitigated", and Dr Neil Pennington
of RWE npower agreed that the delay was "welcome".[98]
British Gas, E.ON and First Utility thought the 2014-2019 timescale
had been achievable, but also saw the change in timescale as "pragmatic".[99]
44. The IET was more cautious about welcoming the
new dates. Dr Thomas told the Committee:
They are better than the old dates, in that they
do give us an additional year to make sure that the specifications
are sound and to fit things in better to the engineering realities.
However, since we don't have the full specifications, we don't
know the details of the bids that have been put in by the DCC
and other communication suppliers, we don't know what their proposals
for assurance will be, we don't know what compromises will come
out of the negotiations over those contracts, therefore we don't
know the full engineering reality of the roll-out of that process.
On that basis, setting timescales now is simply a mistake. At
the very least we need to be flexible, once those things are known,
and to be willing to adjust them again if necessary...timescales
need to be driven by the engineering realities.[100]
Consumer Focus and the Federation of Small Businesses
also thought that there should be sufficient flexibility in the
timescale to ensure that roll-out was done well.[101]
Audrey Gallacher commented:
The way we look at it is that nobody is going
to remember when roll-out commenced or whether it took five or
six years. They are going to remember whether it worked. Let's
not sacrifice what is, after all, a multi-billion-pound programme
for the sake of meeting a date. Let's make sure it fulfils its
objectives in terms of the consumers accessing the benefits of
smart meters as well as industry.[102]
We asked DECC whether there would be flexibility
in the timescale to accommodate any further possible problems.
Baroness Verma said: "I don't see further delays...I think
we are in a very good place now".[103]
45. We welcome DECC's recent announcement that
the dates for mass roll-out are being pushed back by a year. However
there needs to be some flexibility in the new timetable, which
should be driven by engineering and infrastructure requirements
and the need to avoid artificial deadlines acting to push up programme
costs. DECC should be prepared to amend the timetable further
if more time is needed to address any systemic issues that may
arise, to respond to further delays to technical and infrastructure
requirements for roll-out, or to prevent cost escalations for
other reasons.
81 Q 205 [Tony House, Paul Spence, Dr Andrew Pennington
and Andrew Ward]; Q 258 [Stuart Rolland, Darren Braham and Don
Leiper] Back
82
DECC, Written Ministerial Statement by the Rt Hon Edward Davey
MP: Secretary of State for Energy and Climate Change on Smart
Metering, 10 May 2013 Back
83
Ev w5; Ev 80; Ev 87; Ev 99; Ev 126 Back
84
Ev w5 Back
85
Ev 80; Ev 87; Ev 99; Ev 126; Q 220 [Andrew Ward]; Q 247 [Tony
House] Back
86
Ev 87; Ev 99 Back
87
Ev 99. More on EDF's key enablers can be found in its submissions
of written evidence Ev 99 and Ev 104, and in the oral evidence
transcript at Q 211. Back
88
Q 220 [Andrew Ward] Back
89
Ev 65; Ev w75; Ev 80; Ev w27; Ev 126; Ev 146; Ev 150. SMETS 2
= the second version of the Smart Metering Equipment Technical
Specifications. SMETS are discussed in more detail in chapter
5. Back
90
Ev 65; Ev 121; Ev 104; Q 247 [Tony House]; See background section
for more on the DCC. See also Chapter 4, where it is discussed
in more detail. Back
91
Ev 65; Ev 80; Ev 126 Back
92
Q 72 [Allen Creedy]; Q 75 [Audrey Gallacher]; Q 82 [Sean Weir];
Qq 170-73 [Dr Thomas]; Q 220 [Andrew Ward]; Ev 65; Ev w75; Ev
80; Ev 87; Ev 89; Ev 99; Ev w122; Ev 110; Ev 121; Ev 126 Back
93
Ev 80 Back
94
Qq 170-73 Back
95
Qq 170-73 Back
96
Q 220 Back
97
Qq 215-19 [Tony House and Dr Neil Pennington]; Ev 104; Ev 145;
Ev w136 Back
98
Qq 215-19 Back
99
Ev 110; Ev 71; Ev 85; Q 265 [Stuart Rolland, Don Leiper and Darren
Braham] Back
100
Qq 175-77 Back
101
Qq 74-5 [Audrey Gallacher and Allen Creedy] Back
102
Q 69 Back
103
Q 343 Back
| 
