Smart metering - the biggest ever IT project failure?

A NOTE FOR THE PUBLIC ACCOUNTS COMMITTEE

I would like to contribute the following remarks to the committee's inquiry.

1.  Smart metering is the largest engineering project ever undertaken in Europe. Britain alone plans to spend £11 billion, and as I discuss in what follows, I do not believe the costs or timescales to be realistic. Nonetheless £11 billion is almost exactly what the Channel Tunnel cost in today's money. Other Member States also plan substantial expenditures; France's budget (€1 billion for gas an €4 billion for electricity) is less than half ours but still a lot. Overall smart metering is likely to cost households in Europe over €100 billion.

2.  My team has had a research interest in smart grids and smart meters for several years; we've published a number of peer-reviewed papers and have given input to a number of public consultations by DECC, Ofgem and CEER. This work has involved not just lab colleagues but also colleagues at the Foundation for Information Policy Research (www.fipr.org), the UK's leading technology policy think tank.

3.  Our first consultation response, in January 2010, pointed out the lack of evidence that smart meters would save energy, and the lack of any clear specification of what Ofgem was trying to build.[10]

4.  We went into this in much more detail in a further consultation response to Ofgem in September 2010[11] which contains the main substance of our concern with the project. Ofgem were at the time making all the classic mistakes which have been known for years to lead to public-sector IT project failures, and which have been documented extensively in (for example) the case of the London Ambulance Service in the 1990s. As we put it there: "A major procurement is being rushed through for political reasons, with no clear specification, an unrealistic timetable, inadequate financial control, and seriously defective governance; advice on costs and timescales is being ignored; and the proposed architecture ensured continued dominance of metering by energy industry incumbents whose financial interests are in selling more energy rather than less. Pilot projects have failed to show significant energy savings, so even if the systems can be made to work it is quite unclear that hey will achieve anything significant, beyond perhaps a rerun of the Bakersfield fiasco. Finally, the proposed industry architecture stifles innovation and thus guarantees that most of the new, high-value green jobs will go elsewhere".

5.  I commend that paper to the committee and particularly draw your attention to the fact that Ofgem proposed that the DCC should be created in six months - that its contractor could start designing the system in April 2013 and yet by the end of September 2013 have not only designed, developed, tested and fielded a back-end system capable of supporting rich and complex functionality but also implemented and rolled out national communications systems able to support 100 million devices in the field. This is wildly unrealistic; the previous government CIO, John Suffolk, had huge experience with large public-sector projects and took the view that building and deploying any national-scale system takes seven years.

6.  Yet the only material change to the plan since DECC took over the project is that the DCC part is now supposed to take eighteen months, from April 2013 until the end of September 2014. The words "snowflake" and "hell" spring to mind.

7.  The costs currently given are unrealistic. For example, the cost of DCC is not capitalised, but just a notional annual amount for a contract that has not yet been awarded. There are also serious issues about whether it's feasible to fit that many meters by 2019; we don't have enough trained fitters, and the logistics of hiring and training staff, organising meter manufacture and so on are formidable.

8.  We still do not have a stable specification for the meters, or even a workable broad-brush picture. The specification is being worked on by about 20 groups who don't talk to each other and from which experts are often excluded. There is very little technical or project management expertise at DECC. The most critical part of the project - how smart meters will talk to domestic appliances so as to facilitate demand response - is essentially ignored.

9.  There are many issues with the behavioural economics of energy users,[12] the protection of national infrastructure,[13] and the architectures necessary both to protect privacy and promote innovation,[14] about which we're written and in which the committee might be interested.

10.  For present purposes, I predict merely that by the time the next election falls due in 2015, it will be apparent that the smart metering project is just as much a fiasco and a waste of public money as the NHS National Programme for IT was under the previous government.

11.  I therefore urge the committee to call on ministers to either cancel it outright and start again from scratch, or else set clear milestones that are capable of independent verification and which, by advance agreement of Parliament, will lead to the automatic cancellation of the project if and when they are not met.

11 November 2011

11   Consultation response on Smart Metering, Sep 28 2010, at ww.cl.cam.ac.uk/~rja14/Papers/fipr-smartmeters2010.pdf Back

12   On the security economics of electricity metering, at www.cl.cam.ac.uk/~rja14/Papers/meters-weis.pdf Back

13   Who controls the off switch? at www.cl.cam.ac.uk/~rja14/Papers/meters-offswitch.pdf Back

14   Smart meter security: a survey, at www.cl.cam.ac.uk/~rja14/Papers/JSAC-draft.pdf Back