The AMI for the Smart Grid Should Utilise Existing Broadband & Mobile Communication Networks

Sep 16, 2011 No Comments by

Over the past 3 months, an interesting debate has been advancing on the Smart Grid Executive forum on LinkedIn. The question posed was whether an Advanced Metering Infrastructure (AMI) could be delivered by sharing a customer’s broadband connection or if a new bespoke network needs to be developed. The following discussion highlights some of the salient points of the debate, and reviews the main arguments being presented for and against.

Before delving into the detail of the debate, it is worth framing the question in the overall context of the Smart Grid.

A customer based AMI is a key component of the Smart Grid, and its deployment should benefit utilities by improving their efficiency, as well as enabling active participation from electricity consumers. It is likely that the functions which an AMI performs will evolve over time. The AMI will initially allow utilities to dispense with manual meter readings, offer flexible electricity tariffs (including pricing based on two-way power flows) and build a pathway towards demand response. In the medium term, it is hoped that an AMI will enable utilities to obtain detailed visibility of consumer energy consumption and help resolve advanced problems such as power quality. These applications only represent a flavour of an AMI’s functionality, and it is entirely possible that a future AMI may enable applications that cannot be conceived today.

In the UK, the Smart Metering Prospectus (released on 27th July 2010) proposed that the communications system for an AMI will initially be specified by electricity suppliers, before being transferred to a DataComsCo (DCC). The prospectus stipulated that each supplier’s communications system must meet a minimum set of requirements defined in a common technical specification. The prospectus does not mandate any particular communications technology; this will be determined by the market, subject to compliance with the relevant specifications.

The communications infrastructure chosen for an AMI must focus on enabling the applications of the AMI required now and in the near future. The question of whether to build a new bespoke network or rely on existing infrastructure can be tackled through a traditional cost benefit analysis, or more generally by considering the pros and cons of each solution.

Let’s start by considering the advantages and disadvantages of a bespoke network.

A bespoke network could be faster, more secure and more reliable. It could also provide more bandwidth for auxiliary services such as real time power and voltage trace recording. Utilities would be able to assume greater control over their network and if a Smart Grid network roll out was mandated, potentially provide connectivity to more homes than conventional telecoms. A bespoke network also limits supplier risk as the utility does not have to rely on a communications company.

The core disadvantage to a bespoke network is cost. As we know from Smart Grid City, communications deployments can cause costs to escalate dramatically (by 3 times the original estimate in Smart Grid City’s example), and introduce significant delays into the deployment process. EPRI’s AMI analysis in 2007 estimated that communications would make up around 65% of the installation cost of a Smart Meter on a per device basis. It is imperative that Smart Grid projects proceed on time and to budget, or customer and governmental support will wither.

There are also several arguments for and against the shared network.

Perhaps the biggest advantage for the shared network is its comparatively low cost. In contrast to a bespoke network, communications cables already exist to transport broadband traffic and therefore no physical work is required to implement the new network. In the UK, 99.7% of households have access to broadband. Using the existing broadband infrastructure should help to reduce costs and deployment times.

Furthermore, any network performance improvements (for example, bandwidth and latency enhancements) will be subsequently available to AMI applications. Thus, on a shared network, the capacity for AMI applications should scale in line with the expected growth in demand for Internet based services. A shared network solution should also integrate well with 3rd party services and products that might be developed for a customer’s local network.

The biggest disadvantage of a shared network solution centres on security. There are concerns that the shared network will be insecure and an easy route in for hackers.

Furthermore, the technical capabilities of a broadband connection could be insufficient for some Smart Grid applications. For example, if the latency is too high, utilities may not be able to receive real time information or quickly control devices.

On face value, the arguments for and against seem balanced. However, if we take a closer look, the case begins to build against the bespoke network.

A bespoke network could initially be designed with speed and latency advantages over mainstream telecommunications. However, it is likely that this performance differential would be reversed over the medium term. A telecommunications company is driven by its customers to continually invest in new infrastructure to enhance the bandwidth and performance of its service. In contrast, a utility’s metering network would be designed to satisfy the requirements of the Smart Grid now. Further performance enhancements would be made on an ad-hoc basis (if at all) and only when investment was available. The significant danger is that a bespoke network will not grow in capacity over time thus limiting the possibility and / or functionality of future Smart Grid applications.

It is also questionable whether a bespoke network will be accessible to more customers than a shared network. There may be locations in the US which have electricity but no broadband, but it is doubtful that there will be enthusiasm from politicians or consumers to connect these consumers to an AMI ahead of enabling broadband access. Broadband is presumably more valuable to a consumer than connection to an AMI. In the UK, 0.3% of homes have no broadband and it would seem pragmatic to suggest that these homes will not initially be connected to an AMI.

Building and running a bespoke will represent a significant operational challenge for utilities. Telecommunications companies have the most experience of running and building communications networks. By running their own network, the utilities will be taking a significant risk that could manifest in deployment delays and cost overruns.

Another potential disadvantage with a bespoke network is that it is likely to be delivered as a fixed format and size. An insight into the disadvantages of this approach can be gained by considering the SCADA network. The SCADA network was designed to transmit kilo bits over the existing power infrastructure. If SCADA data rates had expanded in line with capacity improvements to the broadband infrastructure, the bandwidth available to the SCADA network would be over one thousand times higher today.

We can also look in more detail into the issue of security implications of a shared network.
In a report published in October 2010, a consortium of companies including Detica, BT and Arquiva highlighted security concerns about the UK’s approach to data security ahead of a DataCommsCo (DCC) being established. They recommended that by delaying the transition to a centrally controlled network and allowing suppliers to use a variety of communications suppliers including mobile and broadband will be a security risk.

It is also worth considering the many examples of online services which successfully secure important information on the Internet. RSA encryption is relied upon by governments, banks and retailers to send highly sensitive information on a daily basis. The sensitivity of the information being transferred across an AMI network is (at least initially) likely to be of comparatively low risk, and comprise of billing information, tariff updates, and technical measurements.

In conclusion, it is my opinion that the case for the shared network is compelling, and warrants further investigation. If the challenges regarding security can be overcome, the Smart Grid has a lot to gain from using a developed communications infrastructure.

Acknowledgement:Thanks to the Smart Grid Executive forum on LinkedIn for inspiring this article and some interesting and enlightened perspectives on this topic.

Government and Energy Policy, Group Discussions on SmartGrid, Latest Smart Grid News, Thinking Grids

About the author

The owner of Thinking Grids is a published author in smart grid topics ranging from smart monitoring and advanced computational techniques for distribution networks, power quality and stability. He's particularly interested in the business benefits of Smart Grid technology, and the overlap between information technology and electrical engineering.
No Responses to “The AMI for the Smart Grid Should Utilise Existing Broadband & Mobile Communication Networks”

Leave a Reply


6 + three =