White Paper: LTE for utilities - supporting smart grids

supporting smart gridsThe introduction of smart meters and smart grid sensors demands a cost-effective and easily deployed communications solution. Laboratory and field tests have demonstrated that LTE networks successfully meet the technical requirements for smart grid communications.

LTE offers utilities very low latency, high throughput and QoS differentiation in a single radio access technology, as well as the commercial advantages of a global standard with a major proven technology ecosystem.

ericsson White paperUen 284 23-3208 | September 2013

LTE for utilities

LTE for UTiLiTiES • from LinEar grid To SmarT grid 2

From linear grid to smart grid With the innovations in power engineering introduced over the past 100 years, it has been possible to scale up electricity grids and to substantially increase their efficiency. over this time, local grids have grown into regional and state grids, which have been interconnected and turned into giant synchronized national and trans-national grids. at the same time, new ways of managing these complex networks have been developed, and the energy demands of users have increased many times over.

a key technology for energy grids in the 21st century is a pervasive data communications network connected to millions of smart meters and thousands of sensors on the low- and medium-voltage lines, devices and sub-stations. in conjunction with the existing communications networks (SCada) used to manage the high-voltage transmission lines and devices, the resulting grid is called a smart grid.

in most countries, the first stage in establishing a smart grid is the introduction of remotely read “smart meters.” Physically, smart meters are electronic meters that measure electricity usage over specific intervals – typically 15-30 minutes – and are connected via a data communications network to a central meter management facility which reads the meters every few days. as well as monitoring energy consumption and generation, some smart meters can also control energy supply to selected devices and report on the status of the grid and premises to which they are connected.

Chief technology officers (CTos) of electricity utilities are searching for proven, cost-effective communications solutions to provide connectivity to the meters and sensors that must be deployed to build their smart grids. Electricity utilities have therefore started to integrate modern standards-based communications technologies – such as LTE – as one of the building blocks of their smart grids.

Studies carried out in laboratory and field environments have demonstrated that LTE cellular networks successfully meet the technical requirements for smart grid communications. LTE also offers the commercial advantages of using a global standard with a major proven technology ecosystem.

These communication networks may be either private (utility owned), public (operator owned), or hybrid (private virtual network over a public operator network), depending on the regulatory and commercial situation of the utility.

This paper discusses the technical requirements to be met by the underlying smart grid communications network, the ability of LTE to meet these requirements as demonstrated in laboratory and field studies, and some specific issues that utilities need to consider when using LTE [2].

GenerationGeneration TransmissionTransmission DistributionDistribution

GenerationGeneration TransmissionTransmission DistributionDistribution

One energy flow, one information flowOne energy flow, one information flow

Bi-directional energy flow, “n” information flowsBi-directional energy flow, “n” information flows

CustomerCustomer

MarketsMarkets

OperationsOperationsServiceproviderServiceprovider

Secure communication flowsElectrical flowsDomain

Secure communication flowsElectrical flowsDomain

CustomerCustomer

figure 1: from linear grid to smart grid [1]

LTE for UTiLiTiES • UTiLiTy CommUniCaTionS 3

utility communicationsmany disparate technologies have been used for smart grid communications, each with their own advantages and disadvantages. Examples include Power Line Communications (PLC), broadband over power line (BPL), optical fiber, mesh radio, WimaX, and 2g and 3g cellular networking. What is evident is that there is no single communications technology that can be technically or economically implemented to support the entire smart grid. Therefore, smart grid communications will have to form a seamless network despite being carried over different technologies.

appropriate technologies must also be long-lived and cost-effective. LTE is a technology governed by a global standard built on mature radio access technology and delivered through a massive global ecosystem. as part of the 3gPP line of mobile networking, LTE shares the heritage of previous generations of mobile technology. Even though the functionality and capabilities of LTE will continue to evolve, the technology platform is mature and stable – so much so that the latest LTE devices are backward compatible with earlier 3g networks (and in some cases 2g networks). Therefore investments made today in LTE technology can be considered future proof.

industry standards There are many industry standards covering different aspects of communications in a smart grid [3].

Communications between the grid operations center and power substations are effectively (but expensively) implemented with SCada networks over optical cables.

However, such solutions are neither necessary nor affordable on the grid between the distribution substation and customer premises; that is, over the medium-voltage and low-voltage networks. Lacking a pervasive communications network, these parts of the distribution grid – despite containing the great majority of grid assets – are typically without real-time management and thus “invisible” to the grid operators.

The introduction of smart meters and grid sensors demands a cost-effective and easily deployed communications solution. This is where wireless communication has many advantages over other communication technologies.

in the past, proprietary communication solutions prevailed in the utility industry. However, with the maturing of iP communications, the trend is towards the carriage of iP communications on standards-based networking technologies.

in terms of electricity industry standards, the iEC 61850 standard is of particular note. originally defined to cover the stringent requirements for automation within the substation, iEC 61850 is proving to be a versatile standard that can also be applied to the medium- and low-voltage networks. iEC 61850 effectively sets the bar for the most demanding communications use case on the medium- and low-voltage grid: notably distribution network automation, which allows the rapid detection and restoration of electricity supply by containing and by-passing faults.

Transmissionsystem

Transmissionsystem

Subtransmissionnetwork

Subtransmissionnetwork

Bulk supplysubstationBulk supplysubstation

Zonesubstation

Zonesubstation

Zonesubstation

Zonesubstation

Zonesubstation

Zonesubstation

DistributionsubstationDistributionsubstation

Servicewire

Servicewire

LV feeder(240/415VLV feeder(240/415V

MV feeder (11kV)MV feeder (11kV)

CustomerpremisesCustomerpremises

Note: voltages will vary with countryNote: voltages will vary with country

LTELTE

figure 2: LTE for distribution grid Communications [4]

LTE for UTiLiTiES • UTiLiTy CommUniCaTionS 4

table 1: message type and performance class from iec 61850 [5]

performance class

requirement description transfer time application

class ms

P1 The total transmission time shall be below the order of a quarter of a cycle (5ms for 50Hz, 4ms for 60Hz).

TT6 ≤ 3 Trips, blockings

P2 The total transmission time shall be in the order of half a cycle (10ms for 50Hz, 8ms for 60Hz).

TT5 ≤ 10 releases, status changes

P3 The total transmission time shall be of the order of one cycle (20ms for 50Hz, 17ms for 60Hz).

TT4 ≤ 20 fast automatic interactions

P4 The transfer time for automation functions is less demanding than protection type messages (trip, block, release, critical status change) but more demanding than operator actions

TT3 ≤ 100

Slow automatic interactions

P5 The total transmission time shall be half the operator response time of ≥ 1s regarding event and response (bidirectional)

TT2 ≤ 500

operator commands

P6 The total transmission time shall be in line with the operator response time of ≥ 1s regarding unidirectional events

TT1 ≤ 1000

Events, alarms

iEC 61850 proposes target values for “transfer time,” that is the time taken for a “message” to be generated, sent and received by grid devices across the communications network. a portion of the transfer time is the time taken for the message to transit the communications network, which would typically be called “network latency.”

The actual time required to execute a standard activity (or use case) depends on the specifics of the use case and the state of the grid and communications network at the time the activity is initiated. While the transfer times are only broadly representative of the timing requirements of a category (“performance class”) of activity, the targets are nonetheless quite helpful when analyzing which categories of activity are feasible for a particular communications network technology.

network latency is thus a key requirement. figure 3 shows the latency measured on operator LTE networks showing round-trip times (rTTs) for different packet sizes. These LTE networks were configured for conventional mobile broadband traffic with no optimization for machine-to-machine (m2m) traffic.

as shown in Table 1, the latency measured across these is low enough to support a wide variety of tasks with performance class P5-P6, including event/alarm notification, handling operator commands, and P4 which covers many slow automatic interactions.

in terms of fast automatic interaction, class P3, Ericsson laboratory tests have demonstrated the successful handling of substation automation tasks with gooSE (generic object oriented Substation Events) messages being carried between grid devices over commercial LTE networks [5]. However, such a demanding activity required careful attention to planning the interaction of devices and particularly the device states. This paper focuses on the use of LTE for mV (medium voltage) and LV (low voltage) devices and meters.

Very low-latency tasks in performance classes P1-P2, which are typically

00 5050 100100 200200 500500 10001000

Mean RTT (ms)Mean RTT (ms)

Length of data packets (bytes)Length of data packets (bytes)

00

1010

2020

3030

Operator 1Operator 2Operator 1Operator 2

figure 3: Latency evaluation in operational LTE networks [6]

LTE for UTiLiTiES • UTiLiTy CommUniCaTionS 5

categorized as “protection” rather than “automation,” are relevant to activities in a substation, between zone substations and on the transmission grid. However, they are not relevant to tasks on the medium- and low-voltage distribution networks.

Both laboratory and field research confirm that LTE is well suited to field area network (fan) communication to distributed devices, or in other words, “last-mile” connectivity. Use cases for electricity distribution include smart metering, distribution monitoring and control, field workforce, distributed energy, micro-grid operation and electric vehicle charging.

technical challengesa study by the Utilities Telecom Council, a global association that promotes the communications needs of electricity, gas, water and pipeline utilities, identified six key technical factors for utility communications: reliability, latency, throughput (capacity or bandwidth), coverage, security and back-up power [7].

how well does lte deal with these factors? > reliability: despite the fact that LTE is a relatively new technology, there are millions of LTE

devices in use around the world today. The commercial success of these services is testimony to the reliability of the networks to be found in any number of locations and conditions. in part, this reliability stems from the use of multiple antenna configurations at both the transmit and receive terminals as a way to deal with radio noise and provide diversity and signal redundancy [2].

> latency: field results, on a dedicated LTE network, without resource scheduling optimization, show that over a range of packet sizes and noise levels, average LTE latency is between 30ms and 40ms. Using iEC 61850 as a guideline, LTE can therefore meet the requirement for many slow automatic interactions at ≤100ms. With network and device optimization LTE may meet the requirement for fast automation interactions at ≤20ms. it is not likely that LTE will meet the most stringent requirement of ≤10ms for protection messages in the near future. However these messages are transmitted in and between zone substations using the optical fiber network and are not relevant to the distribution network.

> throughput: in field tests using conventional end-user devices (dongles), the stationary single

user throughput has been confirmed with latency, scheduler, basic QoS and link budget testing. End devices can be expected to achieve average throughputs of many megabits per second, with the actual instantaneous throughput, determined by radio conditions. While meter reading requires only modest throughput there are some use cases, such as software updates for meters, that could require considerable throughput.

GPRSRel 97GPRSRel 97

EDGERel 99EDGERel 99

EDGERel 4

EDGERel 4

WCDMARel 99

WCDMARel 99

EvolvedEDGE

EvolvedEDGE

HSDPAHSDPA HSPAHSPA LTELTE

Latency (ms)Latency (ms)

00

100100

200200

400400

300300

700700

600600

500500

figure 4: Latency evaluation in operational LTE networks [6]

LTE for UTiLiTiES • UTiLiTy CommUniCaTionS 6

> coverage: Base station cell size varies based on several design factors, notably frequency, antenna configuration, transmit and receive power, and the number of simultaneous users. Some base stations available today also support a maximum cell range feature, so the cell range can be configured. This is particularly suitable for providing coverage in sparsely populated areas where capacity requirements are low. This feature in Ericsson base stations makes it possible to set the cell range between 1km and 100km. Without this feature the maximum cell range is 15km [8]. LTE is designed to support more communicating devices within a cell than previous mobile network technologies.

> security: LTE devices use the same security approach as previous 2g and 3g networks, based on an operator-provided Sim. The security mechanisms used in mobile phone networks have been thoroughly tested over the past decade and have proven very reliable.

> back-up power: The provision of backup power at an LTE base station is a straightforward engineering activity. The capacity of the backup power is readily calculated and provisioned on a private LTE network. When a shared public network is being used by the utility, it may be desirable for the utility to negotiate with the operator for a contract for the supply of adequate backup power.

LTE can be shown to address the technical requirements of utility communications. However, the story does not end there. optimization of the network is needed to efficiently manage the coexistence of different data flows across the network.

m2m traffic models differ from the average phone and mobile broadband traffic models. Knowledge of LTE network optimization through application of LTE features for utility-specific applications beyond simple meter reading is needed to maximize grid efficiency and maintain reliability.

2G2G 3G3G 4G4G

GPRS1998

GPRS1998

40Kbps40Kbps 384Kbps384Kbps 14Mbps14Mbps 42Mbps42Mbps

300Mbps300Mbps

>1Gbps>1Gbps

WCDMA2002

WCDMA2002

HSPA2005

HSPA2005

HSPA+2009

HSPA+2009

LTE2010LTE2010

LTE-A2014

LTE-A2014

figure 5: Peak rate

LTE for UTiLiTiES • PriVaTE or PUBLiC? 7

Private or public?for many industry observers, the next topic for discussion then becomes the question of whether a utility should own a private LTE network dedicated to smart-grid communications, or rent an appropriate operator cellular network. in reality, this question of private or public communications is generally resolved by non-technical realities. for example:

> are utilities able to own spectrum or operate a communications network under the prevailing regulations? is appropriate spectrum available and affordable?

> are utilities able to claim investments in a communications network as part of the “regulated asset base” used by the regulator to determine the prices they can charge wholesale and retail customers?

> is the communications facility to be limited to meter reading, or used for a wider range of applications (distribution automation, field force communications and so on)?

> is the utility requirement “greenfield” (introducing a completely new smart-grid communications network), or “brownfield” (extending the boundaries of an existing legacy WimaX/mesh-radio/PLC network)?

in short, utilities in many countries do not have a choice between private or public LTE as local energy and telecommunications regulations resolve the matter.

There is also the option of utilities using “private virtual networks” carried as a separate logical network by a public LTE network. This middle ground between public and private networks provides some separation of utility traffic from conventional mobile broadband traffic while facilitating simpler charging and support processes.

case study: private lte networka small private LTE network was established for an electricity utility in 2012 as a “proof of concept trial” for field testing and evaluation. The purpose of the evaluation was to test and verify the capabilities of LTE to meet the future communication requirements for last-mile devices, as defined in a set of smart grid use-cases.

Test cases were planned and executed to verify the performance of the test network regarding latency (when stationary), traffic scheduling (when stationary), QoS, link budget, cell range (in normal and extended mode), intra-cell mobility, interference rejection, throughput and performance (high/low noise-level scenarios).

LTE is different from other wireless technologies that have been used by some utilities in that it supports mobile devices. This makes it suited not only to smart metering connectivity, and smart grid in general, but also to mobile use by utilities – for example, for general purpose communications with field crews. The trial network was also tested with these services in mind.

The results showed that LTE meets and exceeds the requirements for fan communications or “last-mile” for smart grid communications.

The customer accepted the results of the testing, their requirements having been met.

case study: shared public networka large public 3g network run by a national operator is being shared by an electricity distribution utility. The utility had previously set up a private WimaX network for meter reading, but found the costs involved in extending the network into lightly populated areas would be excessive.

The prime integrator for the project worked with the meter manufacturer and third-party software and hardware providers to produce a solution that was secure, cost-effective, and integrated with the existing systems already used for the legacy WimaX network.

The utility reads more than 100,000 meters per day over the 3g network.While the utility uses the network of a specific operator, the solution developed does not require

any specific network features from that operator. in other words, the utility is not “locked in” to a specific operator for technical reasons.

for optimal performance, however, it is best for the public network to be configured to meet the needs of the grid communications traffic. for example, grid automation traffic would be

LTE for UTiLiTiES • PriVaTE or PUBLiC? 8

prioritized over broadband subscriber traffic, while broadband subscriber traffic would probably be assigned a higher priority than meter reading traffic. While LTE supports the sophisticated QoS features required, the design of appropriate prioritization requires a sophisticated understanding of the interaction between the grid and communications networks.

LTE for UTiLiTiES • UTiLiTy ConCErnS 9

While the basic technology of 3g/LTE has been shown to satisfactorily meet the needs of utilities, there are concerns that relate more to the LTE ecosystem that must also be considered if a utility is to confidently adopt the technology.

future proofingfew countries today have widespread LTE coverage. as a result, LTE coverage typically exists within a sea of 3g, though this will change rapidly in coming years, becoming at least as pervasive as 3g today. This means that existing networks being used for meter reading are 3g (or in some countries 2g/gSm) today. figure 6 shows forecasts for wireless technologies used by operator networks with 3g/HSPa and LTE growing strongly from 2014, as competing technologies decline.

The LTE standards include two forms of LTE: LTE fdd (frequency division duplex) and LTE Tdd (time division duplex). The fdd form is the more popular with traditional telco operators, who have been planning their investments in spectrum for many years, while the Tdd form is becoming popular with smaller operators (and non-telco users), which have less opportunity to secure the necessary paired spectrum for fdd.

While there is some difference in latency between the two forms, the difference is not likely to be of concern for smart meter communications and distribution automation.

securityUtilities that have built in-house enterprise-style communication networks sometimes assume that “their” network will be more secure than any technology used for public networks.

This assumption is easily disproved. While conventional enterprise networks are hacked on a daily basis, 3g/LTE technology has been designed to be very secure, and the success of this design has been demonstrated in public telecoms networks around the world.

Sometimes the use of “proprietary” networking technologies is actually advanced as a

Utility concerns

Mobile subscriptions(million)Mobile subscriptions(million)

5,0005,000

6,0006,000

7,0007,000

8,0008,000

9,0009,000

10,00010,000

4,0004,000

3,0003,000

2,0002,000

1,0001,000

0020092009 20102010 20112011 20122012 20132013 20142014 20152015 20162016 20172017 20182018

LTEWCDMA/HSPAGSM/EDGE-onlyTD-SCDMACDMAOther

LTEWCDMA/HSPAGSM/EDGE-onlyTD-SCDMACDMAOther

Source: Ericsson (June 2013)Source: Ericsson (June 2013)

figure 6: mobile subscriptions by technology, 2009-2018 [9]

LTE for UTiLiTiES • UTiLiTy ConCErnS 10

way of implementing security. This is a dubious argument, as it is not in the interest of the proprietary vendor to be transparent about the real success (or otherwise) their equipment has had in defending against attacks.

Whether utilities use public or private LTE networks, the security engineered into networks is robust and proven, with the security models used by operators reporting good success.

subscription managementThe communications modules used with smart meters are uniquely identified, not just with the module serial number but also with a mobile network device identifier contained in a Sim. mobile phone users know them as the chips mounted on tiny plastic cards that are provided by the mobile phone operator when they have signed up for a new phone package.

a new business process is required to allow the modules and associated Sims to be purchased, delivered and tracked to the point of installation. The management of the modules must continue throughout the life of the communications service, up to the eventual termination of the service.

This activity is important when providing a wide range of m2m services through LTE networks. as a result, there are a range of proven systems that are available specifically to meet this requirement.

device management The deployment and management of large numbers of smart meters and smart grid devices introduces a number of challenges for utilities. manual processes and human interaction involved in managing device volumes are not viable when the number of endpoints extends to the thousands, hundreds of thousands, or millions.

Utilities need technology to automatically process device management transactions over the lifecycle of a device. for example, when initially deploying a device, utilities expect that a device is able to automatically attach to the desired network and authenticate seamlessly with server-side software used to communicate with it. Similarly, when updates to firmware or configuration parameters, utilities need systems that can manage these instructions with minimal manual intervention.

Utilities are increasingly demanding standards that assist with device management and are enabled in communications modules being developed by players in the industry. at the forefront is the open mobile alliance (oma), which has created a set of widely used standards for both the provisioning phase of a device and remote configuration of communication parameters along a device’s lifecycle.

spectrumLTE networks use licensed spectrum, which in most countries is licensed by the government to the highest bidder.

assuming the regulators allow utilities to license the use of appropriate spectrum, the cost of spectrum is a new item for utilities to deal with; it may not be considered part of a utility’s regulated asset base, which determines the costs it may recover in user charges.

The situation varies considerably from country to country, and local regulatory rulings must be taken into account.

The use of unlicensed spectrum (often used by mesh radio communications) is not attractive for anything but “best effort” communications carrying the most basic meter reading, which can tolerate the slow and erratic throughput of an unplanned radio environment. for example, this approach would not be satisfactory for even slow distribution automation activities (as specified in iEC 61850).

LTE for UTiLiTiES • ConCLUSion 11

conclusionLTE offers very low latency, high throughput and QoS differentiation in a single radio access technology that is supported by a global standard. The evolution of LTE through global standardization ensures a future-proof technology that does not compromise investments in network infrastructure.

Laboratory and field tests show that, as a communications technology, LTE meets and exceeds the requirements for fan communications, or “last mile” for smart grid communications.

for utility-specific applications beyond simple meter reading, optimizing the LTE network through application of LTE features is needed to efficiently manage the coexistence of different “use cases” on the network, and to maximize network efficiency and maintain reliability.

in short, the design and configuration of an LTE network to meet the requirements of a distribution utility requires detailed understanding of the communications technology, and of the consequences of design and configuration choices on the communications network and thus – most importantly – on the actual operation and management of the electricity grid.

When utility CTos and technology strategists are planning the use of LTE as a data communications solution within their smart grid – whether using a private LTE network or a shared public network – they tend to find it is prudent to work with LTE specialist suppliers and integrators to ensure targets for both communications and energy networks are met.

finally, it is interesting to note that the communication needs of other utilities, such as gas, heat and water utilities, are very similar to those of electricity utilities, and similar solutions are already being planned and deployed by these utilities. This raises the prospect of common communication networks being used for several utilities with overlapping territories.

it is clear that LTE is beginning to play an important role in the day-to-day business of utilities.

LTE for UTiLiTiES • gLoSSary 12

GLOSSARYbpl broadband over power linefan field area networkgoose generic object oriented Substation Eventsiec international Electrotechnical Commissionlte fdd LTE frequency division duplexlte tdd LTE time division duplexlv low voltagem2m machine-to-machinemv medium voltageplc Power Line CommunicationsQos quality of servicertt round-trip timescada Supervisory Control and data acquisition sim subscriber identity module

LTE for UTiLiTiES • rEfErEnCES 13

References1. niST Smart grid framework 1.0, September 2009.

2. E. dahlman, S. Parkvall, J. Sköld, 2011. 4g LTE/LTE-advanced for mobile Broadband. iSBn: 978-0-12-385489-6. academic Press, oxford, UK.

3. international Electrotechnical Commission (iEC) 2012. Smart grid: Core iEC Standards, http://www.iec.ch/smartgrid/standards/

4. n. Higgins, V. Vyatkin, n. nair and K. Schwarz. distributed Power System automation With iEC 61850, iEC 61499, and intelligent Control. iEEE Transactions on Systems, man, and Cybernetics – Part C. applications and reviews, Vol. 41, p.81-92, no. 1, January 2011

5. iEC 61850-5 ed2. Communication networks and systems in substations – Part 5: Communication requirements for functions and device models. 2003-07. 61850-5 iEC:2003(E).

6. y. Xu, Latency and Bandwidth analysis of LTE for a Smart grid. master Thesis, royal institute of Technology, Stockholm, Sweden. Xr-EE-rT 2011:018. 2011.

7. Utilities Telecom Council, “a Study of Utility Communication needs: Key factors that impact Utility Communication networks,” Sep 2010.

8. Ericsson. User description maximum Cell range. document: 67/1553-HSC 105 50/1-V1 Uen d.

9. Ericsson mobility report, June 2013

© 2013 Ericsson aB – all rights reserved