Energy and Climate Change CommitteeWritten evidence submitted by John Cowburn (SMR01)

The following comments are made under the headings provided by the Energy & Climate Change Committee and are based on 28 years experience in the metering industry in the design of electricity and gas meters at Siemens, Landis+Gyr, PRI and Bglobal Metering. The comments relating to standards are based on 20 years involvement in UK, European and International standards bodies including as Chairman of ZigBee EURO SIG, lead UK delegate to CENELEC TC13 WG2 (Smart Metering Standards) and convener of IEC TC13 WG15 (Payment Metering Systems).

Are the Government’s cost and timescale predictions for roll-out realistic and will it deliver value for money?

The timescale for roll out will be severely affected by the lack of clarity in the specifications (SMETS and the DDS’s). The specifications have failed to take cognisance of European Standards with many of the requirements being ambiguous and in conflict with the recognised harmonised standards and not aligned with recommendations made by the European Smart Metering Coordination Group. This may be cause issues with the MID as the meters will be “country specific”.

The specifications for the meters have grown to include many features not originally envisaged. These extra features have had a knock-on effect on the communications requirements. The specification for the ZigBee HAN has increased from 350 pages to 593. Every message packet must be tested in every meter with every other device in the network in order to ensure interoperability requirements have been satisfied prior to release of the final specifications. Every new feature introduces a new set of tests and the time to certify the specification grows exponentially as a result. The ZigBee specification has now grown to such an extent that it will be impossible to test all the features in the time available.

The specifications have been developed on the assumption that the communications will fail, requiring all meters to work with and without communications. The additional functionality is built into every meter at additional cost and complexity. The majority of the complexity is to cater for prepay working when communications fail whilst ensuring there is no room for fraud. All meters are fitted with this even though it is only used by 20% of the population and it is only a small percentage of these that will fail and an even smaller percentage that will suffer from fraud attempts. 100% of the meters are loaded with extra cost to cater for less than 5% that will need it.

In addition, DECC has imposed new requirements for end-to-end security which has further knock-on effects on the DLMS/COSEM and ZigBee specifications that will add further delay to the programme—well over 12 months. (Probably two years.)

In terms of value for money, the cost of the metering has increased from the impact analysis as the specifications have grown. The assumed cost of the metering used for the impact assessment was £139 but this has increased significantly and is likely to exceed £300 including installation cost. The power consumption taken by the metering exceeds that taken by present day meters, costing £5/year. When the cost of the DCC is added, the annual cost per home is likely to be in the region of £40/year. Consumer Focus claims that the average household saving will be £30/year so it is difficult to see how the metering is value for money and the figures resulting from the impact assessment are no in doubt.

A review of the SMETS2 specification revealed that added costs introduced due to the metering architecture based on the “Thick Meter/Thin Comms Hub” approach increased the cost of the metering by £1.7 billion for the roll out. It was also noted that the cost to the country to power 27 million IHDs at 0.6W would be £200 million over 10 years assuming electricity costing 15p/kWh.

The DECC architecture specifies a smart gas meter and a gas meter mirror in the communications hub. Every home will effectively be fitted with two gas meters. If the gas meter were a simple device and the smart gas meter functions were in the mirror then the cost could be reduced by approximately £30/home.

What are the potential benefits of smart meters for consumers, and what barriers need to be overcome in order for consumers to realise them?

The metering system as specified assumes savings can be made by showing the consumer their energy usage in terms of kWh and in monetary terms. One question is—what is the purpose of smart metering? Is it to save money or to save energy? If the latter then smart metering can show very well how much was used and when, the former is not quite so simple as cost changes with time of use and/or quantity used, a fact that DECC has failed to grasp.

When showing past consumption data in money terms, the IHD as specified does not show how the cost was affected by tariffs. The cost can vary with time—up to 48 rates applied according to a switching matrix containing up to 200 switching times that can be time of day, day of week, seasons and holidays. Cost can also vary with usage (block tariffs) that change according to billing periods so energy may be charged at up to four different rates depending on previous use. Just to add to the complexity, the cost can vary with both time of use and with blocks making it virtually impossible to work out the cost at any point in time.

It is impossible to determine what behaviour lead to the historical cost of energy used.

It is also impossible to plan ahead since the IHD as specified does not inform the consumer what the forthcoming rates will be. The consumer is not told that energy will be cheaper later on in the day.

Potential benefits for consumers with smart meters would come from a simpler architecture where a simple meter supplies metering data to a home hub that is supplied with tariff data from the supplier. The hub could then provide data to smart appliances within the home to automatically choose the best time to run and choose the most appropriate fuel to use. Such an architecture would not only be lower cost than the DECC specification but it would also provide smart grid functionality. This architecture has been shown to work in Korea for example.

Is there a possibility that suppliers will gain considerably more than consumers from smart meters? Is enough being done to ensure that any financial benefits accruing to suppliers will be passed on to consumers?

It may well be that the suppliers find themselves having to provide call centres to help customers with IHD installation and fault diagnosis since these devices are controlled by them. The suppliers don’t seem to have realised that with the architecture as specified by DECC they will be required to control the joining of IHDs onto the HAN. The amount of data returned from the smart meters is huge and suppliers may find themselves inundated with data they can’t handle (there are 512 items in the event log for example for each meter). Costs for suppliers may therefore increase.

What lessons can be learned from successful smart meter implementation and usage elsewhere in the world?

The Dutch experienced issues with privacy which took two years to be addressed and overcome. They have also addressed the issue fitting a switch in every meter, determining that the risks to the national infrastructure exceed the benefits and they removed the requirement, removing the risk and reducing the cost.

The Koreans have deployed smart meters with smart hubs. The system provides automated control of appliances to provide smart grid functionality whilst maintaining comfort to the consumer. This has been achieved using ZigBee 1.0 from 2006, without resorting to the complexity of the SSWG proposed extensions.

Several utilities in the USA have successfully rolled out millions of ZigBee 1.0 with simple meters, in-home displays and smart controls.

British Columbia Hydro has begun a roll out with ZigBee 1.1 with a simple block tariff and have implemented a head-end based prepay solution that does not require the complexity of a prepay feature within the meters.

The Australian state of Victoria is installing ZigBee 1.0 based meters throughout the state to work with IHD’s and water meters.

The common thread running through the successful implementations is the use of a simple meter with a simple implementation of the HAN to provide a reliable and easy to test solution that is easy to maintain. The “smart” functionality comes through the other devices in the system at the head end and within the home. Interoperability can be easily certified with limited functionality and reliability can be assured.

Will smart meters empower customers to take greater control of their energy consumption?

See earlier comments on the shortcomings of the IHD. Customers can neither see how their costs were affected by usage time or by block, neither are they provided with the data to plan ahead. It is therefore difficult to see how consumers can analyse their consumption or plan ahead using the IHD. It can only work if the tariffs are simple and easy to understand.

Automation is the only the route to energy saving unless the tariffs are very simple and easy to understand with a direct relationship between energy used and cost.

Will consumers on pre-pay meters obtain the same benefits from smart meters as other consumers?

All smart meters will be capable of prepay—they will all be fitted with switches/valves adding £5 per meter, £10 per dual fuel home. 80% of consumers are not on prepay. 21 million homes will be fitted with a cut-off switch that will not be used at a cost of £210 million. Credit customers will have over specified meters fitted at extra cost to provide a “level playing field”. Unfortunately, the IHD will not provide prepay facilities so a prepay consumer will require an “enhanced” version. This will cost approximately £40 so the cost to serve prepay customers will still exceed that of a credit customer.

Prepay customers will have the same benefits as others but the others will be bearing excess cost to support features that are not required.

It should also be noted that Smart Meters would present users with some major issues should the communications suffer failure. They would be faced with inputting a 20 digit code through two push buttons on most meters to load the credit. On average this would require 120 button presses without making a single mistake, not easy for even the most dextrous person—and this may have to be performed in the dark. This compares with the simple insertion of a key today, a simple and quick operation.

Should vulnerable customers and the fuel-poor be first in line for smart meters so they can get the benefits sooner?

The fuel poor would gain far more benefit from modern programmable thermostats set up correctly. Such devices cost as little as £30 and can save £100 per year with no loss of comfort. The data provided on the IHD is unlikely to be of any use to the fuel poor as it will be impossible to make much sense of the data presented.

Vulnerable customers—eg those suffering from dementia—could find themselves being cut off through non-payment of bills through forgetfulness (based on personal experience with BT).

What is the best way of involving third-party trusted messengers, such as charities, consumer groups, community organisations, local authorities and housing associations in roll-out?

The message is going to be difficult to sell based on the excessive cost of the metering compared to savings they will generate. Consumers must want the meters, they will not accept them being pushed, particularly in the age of austerity.

What are the potential obstacles to rolling out smart meters in the UK and how should these be addressed? What pitfalls have hindered roll-out programmes elsewhere and are we doing all we can to avoid them?

One massive obstacle is that of the switch fitted in every meter. Today only 20% of homes are fitted with meters with switches for prepay. Today, power is switched to local areas through a very expensive and highly reliable breaker in the substation often remotely controlled, serviced regularly and monitored. A smart meter by comparison is fitted with a low cost switch controlled through software running on a low cost microprocessor with little or no fault tolerance built in. The meter is never serviced and is expected to work reliably for 20 years. Smart meters will therefore introduce a new failure mode that may result in power cuts. These power cuts can only be rectified by visiting each house and replacing the meter, an impossible task to complete quickly if it occurs on a large scale. It is possible that large populations may well remain without power for several weeks should a batch of meters suffer catastrophic failure.

Reliability of switches in meters has not been good over the last 30 years. Large scale switch failure has occurred in Key Prepay Meters in Southern Electricity (100,000—the whole of the installed base), Radio Telemeters (all of South Wales failed on a single winters night), SME Smart meters currently installed and a batch of Quantum Prepay Gas Meters.

DECC has failed to specify a reliability requirement, current smart meters fitted with switches are failing at a rate of 3/week in a population of 140,000 which was deemed acceptable by the AMO. This would result in a failure rate of 580/week if the same failure rate was experienced in the roll out. This would clearly be unacceptable. If DECC were to specify a failure rate of less than 1/100,000 per year where customers were left without power then manufacturers would find it impossible to meet this but it would still mean 270 homes would lose their electricity supply every year.

There is also a considerable security risk in fitting switches/valves in every meter. The potential threat to critical infrastructure from accidental or intentional sabotage attacks cannot be underestimated. A successful attack leading to permanent loss of supply to thousands or millions of homes would lead to civil unrest and death.

The switch in the meter serves no useful purpose to 80% of the population. It does nothing for smart grid functionality as claimed, cutting the supply to a house is not a particularly smart way to control load. The switch is not a safety device and cannot be relied upon to provide isolation so cannot be used to isolate empty properties for example.

Whilst other countries such as the Netherlands have reviewed the inclusion of the switch in smart meters, DECC has so far ignored the issue hoping that additional security measures will solve the problem, which unfortunately it won’t.

Are levels of public awareness of and support for smart meter roll-out increasing?

Public perception appears to be increasing but those who are aware of their shortcomings are not supportive.

Is enough being done to increase consumer awareness about smart meters? Could DECC’s consumer engagement strategy be improved?

Consumers are going to have their energy bills increased by at least £30/year, probably more like £50 to pay for a smart meter that may save them £30 if they are lucky. This follows on top of the increase to pay for the measures in the recent Energy Bill.

DECC may well find a public backlash as experienced in Victoria where the government was brought down partly as a result of the costs associated with smart metering.

DECC may wish to consider the inclusion of the metering cost on the energy bill to show clearly what they are costing, as is the practice in Australia.

Are consumers’ concerns about privacy and health being addressed adequately?

Privacy concerns are being addressed but DECC would be advised to take note of what happened in the Netherlands. The Dutch have offered help in this area but DECC have not taken up the offer.

Is there any evidence that consumers’ concerns about smart meters are declining or growing?

There is not much evidence in either direction.

Will the commercial benefits of smart meter roll-out be captured within the UK?

The vast majority of the meters will be manufactured off shore. Itron have just closed their UK factory and moved production to France and Germany. Landis+Gyr manufacture in Poland and China, Elster in Germany and China, EDMI in Singapore and Malaysia, GE in China, Secure Meters in India. There is some local content with Itron in Manchester (gas meters), and the other major players with some final operations such as calibration and distribution based in the UK.

All the semiconductors are made off shore although there is some UK involvement with Silicon Labs and NXP in Cambridge and Sheffield supplying ZigBee technology for the HAN.

Head end system providers are based in France (Itron), Belgium (Energy ICT), Switzerland (L+G). A few small players are UK based.

WAN providers are based in the USA (Silver Springs and Sensus) with a couple in the UK—Arqiva, BT. GPRS modules are manufactured in China.

The biggest benefactor from smart metering will be China.

Will DECC’s current approach to roll-out, including on procurement and establishment of the central Data and Communications Company, deliver an optimal data and communications strategy?

The Comms Hub will be provided by a Communications Service Provider (CSP). This company will have sole responsibility for designing, building and supplying all communications hubs. It is a monopoly without competition and therefore not subject to price pressure. The arrangement is much the same as that enjoyed by the GPO prior to privatisation, supplying black Bakelite telephones to the whole country. This cannot be a good model to follow in the 21st century.

What criteria should DECC use to measure the ongoing success of roll-out?

Public reaction will be the best measure. If there is acceptance by the public to pay for smart meters then they will have been convinced they have value and there will be market pull. It will be impossible to impose the solution on a reluctant public.

January 2013

Prepared 26th July 2013