JRC Response to the consultation A SMART, FLEXIBLE ENERGY SYSTEM A call for evidence

Demand side management does create a number of scenarios under which it might prejudice the stability of the electrical network.  As communications specialists, JRC’s focus is that it is likely that demand side management will be dependenton telecommunications signals to trigger those responses, and these may be subject to ‘common mode’ failure situations (for example when all public telecommunications networks fail simultaneously due to a wide-area mains powersupply incident).

Common mode failure of the telecoms required to implement demand response could be potentially catastrophic where the cause of the failure – or disruption – of the telecoms networks correlates with stresses to the energy networks.  Conceivable scenarios might include:

• New Year’s Eve, when telecoms networks, especially elements such as the short message text service, become overloaded just before and after midnight and congestion management mechanisms are implemented, but at the same time the load on the electrical power networks can be unpredictable, requiring demand side management for mitigation of unexpected peaks.
• Flooding incidents where telecoms, electricity and gas networks can be degraded, but in unpredictable ways, thereby requiring demand side management of the power network at the same time as telecoms are disrupted.
• Storm conditions when electrical networks are disrupted which could be mitigated by demand side management, but the telecoms networks which would facilitate these measures are also disrupted.

If this is recognised as a potential vulnerability, JRC would be interested in working with the industry to develop options for mitigation of the vulnerability.

The Department of Business, Energy & Industrial Strategy (BEIS), the Office of Gas and Electricity Markets (Ofgem), and the Office of Communications (Ofcom) have differing interpretations of policiesand responsibilities with regards to businesses, citizens, and consumers. Under the Communications Act, Ofcom interpret the obligation to serve the interest of ‘consumer/citizens’ as according higher priority to direct consumer-facing services –mobile data and broadcasting –than communications directly related to delivering cost effective and reliable essential services for the consumer/citizen.

Almost all products and services offered to UK’s businesses, citizens, and consumers rely directly or indirectly on a reliable, affordable and sustainable energy supply.  It would be helpful if the dependence of 21st Century energy networks on telecommunications was recognised, and that consumers and citizens benefit directly from utilities having access to resilient radio communications systems to supervise and control the critical infrastructure electricity and gas grid networks.

Once agreed, it is hoped that the three organisations will recognisethat the move to Smart Grids is only likely to require a modest increase in the existing typical data rates, e.g. from 9.6 kbit/s to 64 / 100 kbit/s, and therefore only a correspondingly modest increase in spectrum access. Indeed, only 2 x 2MHz of additional 400 MHz Band private spectrum is predicted to be required in addition to the currently allocated 2 x 1 MHz of 400 MHz spectrum.

• Whilst this may be seen as a 200% increase, it should be noted that the average additional future private spectrum requirements for Critical Infrastructure Utility Operations Networks is likely to be equivalent to only ~1.5%of the 1,200 MHz of spectrum identified for public mobile / IMT systems.

Answer 34a: the distribution network operators (DNOs) are already aware of the regular peaks and troughs of power demand throughout the average days and, for example, special event days so they can prepare for them well in advance.

The use of 30-minutes time-slot pricing mechanisms may help to control the times that vehicle charging is typically undertaken. It may be, however, that some users are prepared to pay whatever price in order to have the flexibility to connect to the network whenever and wherever they want to. If the power to these areas is uncontrolled, this may cause an issue if enough devices are connected unexpectedly in an area that is already operating close to capacity.

It should be noted, however, that, assuming there is sufficient radio spectrum access, the SCADA systems, etc, should detect these unexpected increases and report them back to the control centre so that additional power may be provided to those areas automatically.

Answer 34b: The batteries in electric vehicles have a limited number of discharge / re-charge cycles before they need to be replaced. It is understood that the cost of replacing the batteries is very high. Once aware, owners may be reluctant to allow their batteries to be used for putting electricity back into homes, businesses, or the network for fear of the high cost of replacing the batteries. Where vehicles batteries are leased rather than bought outright, presumably the leasing company will determine whether they can be used in vehicle-to-grid mode or not asit will potentially degrade the service life.

Electricity and gas networks have traditionally protected their control systems from cyber-attack by the inclusion of an air gap between their resilient telecommunications networks and publically accessible networks, especially the Internet.

In the future, there is concern that the regulatory regime for the energy sector will make the separation of operational and commercial networks financially unattractive, or effectively impossible. Public telecommunications networks generally give indicative service levels, and provide a ‘best endeavours’ service.  This is not suitable for monitoring and controlof critical infrastructure.

At present, smart meters are less sensitive elements of network control, and can therefore be connected with less resilient public mobile networks and licence-exempt radio services, although this may change in the future if they become critical elements in energy network management, as is being considered in some countries.

This consultation is not the place for a detailed analysis of the security vulnerability of ‘smart networks’, but thinking which assumes that all monitoring and control will be conducted through the public internet needs to be challenged.  Although monitoring and control is generally migrating to ‘Internet Protocol’, this does not mean that the equipment has to be connected via the public internet.  Separate and private communications networks are easier to secure than the public internet.  In an environment where hostile ‘actors’, some backed with government resources, see action against energy networks as a way to undermine an economy, separating critical monitoring and control networks from public networks may be a useful counter measure.  [Separacy of itself is not adequate as was illustrated by the ‘Stuxnet’ attack on Iranian nuclear facilities, but interconnection of ‘operational telecoms’ with the public internet creates more opportunities for compromise as was demonstrated by the attack on the Ukrainian power network.]

Even where data security has not be compromised, dependency on the public internet has demonstrated at least two vulnerabilities which can be mitigated by separate secure ‘operational telecoms’ networks:

• Denial of service attacks where the public internet get so overloaded than all traffic becomes congested and delayed, even if the contents of the secure messages are not compromised.  This is not critical for most services which can tolerate delays, but can be fatal for real-time control where responses are required within defined time periods, in some case sub-second.
• Lack of identification of separate redundant telecoms routing, such that a single physical failure – the destruction ofa bridge carrying telecoms cables or a fire in a cable tunnel results in interruption to both main and back-up telecoms circuits.

Answer 43a: Efficient use of system resources (including flexible resources and innovative solutions) for local network and system-wide operations: 

Currently, the efficient use of system resources includes the incorporation of resilient private radio SCADA systems to monitor ~10,000 of the UK’s electricity sub-stations. Efficiently will increase significantly because this is expected to rise to ~100,000 as the grid becomes increasingly smart. The number of monitored locations increases further when the distributed generation sources, etc, that require resilient communications are added to this total.

Caution should be observed before assuming that any new ‘innovative’ solutions will always be better than existing solutions. For example, the upload data rate (144kbit/s) of the new innovative 200 kHz, ‘narrow band’ LTE (NB-LTE) systems is lower than the upload data rate (153kbit/s) of a standard 50 kHz private digital system. So, in this case, the narrower channel system is ~4 times more efficient than the 200 kHz wide-band channel system.

Communications technologies installed by the critical infrastructure utilities need to use long-life technologies rather than technologies that have a limited lifetime, e.g. public mobile GSM systems. Indeed, in 1985, Ofcom allocated 2 x 1 MHz of 400 MHz spectrum for the use of SCADA systems within the UK. During these past 30 years, the primary innovative solution that has been introduced into those systems is their changing from 12.5kHz narrow band analogue systems to 12.5 kHz narrow band digital systems. This change enabled an increase in typical data rates from 2.4 kbit/s to 9.6 kbit/s. NB: these SCADA systems now need to be updated to include the ability to operate 64 / 100 kbit/s in 25 kHz channels and, perhaps, 154 kbit/s in 50 kHz channels. The critical infrastructure utilities are therefore seeking 2 x 2 MHz of additional 400 MHz spectrum so that as to have the flexibility to introduce these two options and other wide-band channel options.

It should also be noted that the current narrow band private radio systems, as well as being resilient, operate to the remotest parts of the UK. This is beyond the economic coverage areas of most public mobile systems. Further, the communications infrastructure is in place already so, providing it remains within the current 400 MHz frequency band the cost of the expansion to cover ~10 times more remote sites should be relatively low when compared with changing to a higher frequency band(s).

Answer 43b: Resilient and secure system with efficient emergency and system recovery procedures: 

Whilst the existing and Smart Grid systems may be included under the general heading of Machine to Machine (M2M), they are more appropriately referred to as Resilient M2M (RM2M). This highlights that standard communications systems, including public mobile based M2M, LTE, and IoT solutions, are unlikely to be suitable for critical infrastructure systems. Electricity and gas grid networks are cautious to ensure that their control systems are resilient by, among other measures,the inclusion of an air gap between their networks and publically accessible networks, especially the public internet.

It should be noted that the base stations of critical infrastructure utilities typically have up to 96-hour power backup resilience. This is to enable efficient emergency and system recovery during the most extreme weather, etc, conditions. The cost of this power resilience is high and is therefore not typically included in standard communications systems. It is particularly important that there is resilient wide-area communications to remote rural areas because this is often where damage to the power network is likely to occur during heavy storms.

System resilience is also part of the health & safety aspect of the wide-area instant-access voice communications networks used for day-to-day maintenance and emergency repairs.

Whilst they may not recognise the importance of system resilience, the public are acutely aware of their dependence on affordable, reliable and sustainable energy supplies.

Answer 43c: Responsiveness to and readiness for rapid / unpredictable change: 

The responsiveness to, and readiness for rapid / unpredictable change, is typically managed by the critical infrastructure utilities’ supervision, control, and data acquisition (SCADA) networks. Any changes are reported to a communications hub that addresses the issue automatically. This can be as simple as detecting the loss of power from a faulty sub-station and re-supplying power to that area via an alternative sub-station / route. In this context, it is important to note that radio technology is of particular importanceas it can be deployed more quickly than wired networks, often at lower cost as well.

Answer a: Transmission companies already remotely monitor theirsub-stations using resilient private SCADA communications networks. This remote monitoring is expected to increase to ~100,000 sub-stations. The data acquired from the DNO’s sub-stations, some of which are in theremoterural locations, should be communicated to the transmission companies so that they may plan and control the transmission supplies more efficiently.

• Whilst this increase in communications requirements may be onerous, the UK’s critical infrastructure utility operations have more than 30 years’ experience of designing, installing, operating, and maintaining their own private SCADA / private resilient machine to machine (RM2M) systems so they are well qualified to establish the appropriate communications systems to use for smart grids.

Answer b: The DNOs, etc, would very much like to manage their networks more efficiently. Unfortunately, the lack of suitable radio spectrum is a major contributing factor that is preventing them from achieving this. As highlighted elsewhere in this response, the DNOs are seeking an additional 2 x 2 MHz of 400 MHz Band spectrum. This spectrum will enable them to remotely supervise and control most of their infrastructure. This will enable them to sweat their power assets and quickly identify, and rectify, any faults on the system.

On the communications side, the transmission and distribution operators appear to work very effectively together. They understand that different areas of the country require different communications solutions. For example, there are plenty of opportunities for private fibre communications in Central London whereas this may not be available in other areas. (If mandated, however, the very high cost of fibre, when compared with private radio solutions, would ultimately be passed on to the consumers.)

There is still also an on-going debate about whether the way in which the UKhas implementedsmart meter communications will deliver the connectivity required for real-time control.  In most other countries, distribution companies are responsible for smart meter communications and back-haul data through the local substations, facilitating real-time control and distributed intelligence.  It is not clear at this stage whether the UK’s chosen architectural model will allow this model of distributed control to be adopted.

In following the outcome of LCNF/LCNI projects,most have a telecommunications requirement as an essential element of the project. Frequently this is identified as a weakness, and although the chosen telecoms solution is adequate for the purposes of proving the project concept, it would be inadequate as a long term solution if the project were rolled out into ‘business as usual’ on a wider scale.

In JRC’s view, there is a need to identify the telecoms element of projects more clearly, and trial telecoms solutions being adopted in other countries – mainly radio solutions, but currently not being deployed in the UK. The reason often relates to access to suitable radio spectrum, and the likelihood that ifa wireless technology is adopted, whether adequate radio spectrum at suitable cost would be released by the UK radiocommunications regulatory authorities.

At present, there is an initiative to investigate release or sharing of radio spectrum held by government bodies which is either unused or not fully used.  It is felt that closer co-ordination between the relevant government departments and regulators – BEIS,DCMS, Ofcom, Ofgem and the Central Management Unit could unlock progress to progress trials in this area with the potential for world-class innovation to the ultimate benefit of UK electricity and gas consumers through more affordable, sustainable and secure energy supplies.