Smart Metering Program in India - A Critical Assessment
Smart Metering Program in India - A Critical Assessment
Smart Metering Program in India - A Critical Assessment
India is presently rolling out 250 million smart meters on fast track and can leapfrog to AMI 2.0 by leveraging the
experiences of global utilities who have successfully ascended to AMI 2.0. This paper examines the ongoing AMI
rollout in India and suggests the measures for mid-course correction to protect the investments.
Disclaimer Authors
Anand Singh
The information and opinions of this document
Balasubramanyam K
belongs to India Smart Grid Forum (ISGF). ISGF has no
Disha Khosla
obligation to communicate with all or any reader of
this document when opinions or information in this Rajani A
document changes. We make every effort to use Reena Suri
reliable and comprehensive information, however we Reji Kumar Pillai
do not claim that it is accurate or complete. In no Vivek Gupta
event shall ISGF or its members shall be liable for any Shashi Bala
damages, expenses, loss of data, opportunity, or
profit caused using the contents of this document. October 2023
As of 15 August 2023, AMI projects totaling about 230 million meters have been approved by the nodal
agencies (PFC and REC) and contracts for 56 million meters have been awarded; tenders for the rest are
under various stages of finalization. During the RDSS finalization and program rollout in the past 2 years,
the SBD has gone through several rounds of amendments. Additionally, certain conditions imposed on the
project implementation methodology do not align with successful practices from smart metering
experiences worldwide over the past two decades. This paper provides insights gained from experts with
decades of hands-on experience in implementing and maintaining large AMI systems in different utilities
around the globe. It also offers a practical implementation roadmap for smart metering in Discoms in India.
Some of the key components of the AMI systems and the considerations for their selection and
implementation are discussed here.
Smart meters are a significant improvement over traditional meters in many ways. They offer several unique
characteristics that can greatly benefit both consumers and utilities. One key feature of smart meters is their
ability to transmit detailed and accurate energy usage data in real-time. This enables consumers to monitor
and adjust their energy consumption patterns, leading to greater energy-efficiency and cost-saving benefits
over time. Additionally, smart meters have the potential to support a more reliable and efficient power grid
owing to their ability to remotely detect and report power outages and other issues. Smart meters eliminate
the need for manual meter readings which saves time, reduces costs and labor, and improves accuracy. With
benefits ranging from improved energy-efficiency, reduced costs, to better monitoring and reporting
capabilities, smart meters are a valuable upgrade for any utility looking to improve their services.
India is one of the few countries that have a national-standard for smart meters. IS:16444 standard for smart
meters was issued by Bureau of Indian Standards (BIS) in 2015 and the associated data communication
standard IS:15959 Part-2 was issued in 2016. Presently there are 87 BIS certified meter manufacturing units
with cumulative annual capacity of over 100 million meters in India3. While new domestic companies are
setting up manufacturing facilities, some of the existing players are augmenting their manufacturing
capacities as well. Foreign players are not expected to jump in due to very competitive pricing by domestic
players and ban of import of smart meters from countries sharing land border with India (which precludes
Chinese firms from participating in the smart metering projects in India). Overall, availability of smart meters
is not expected to be a constraint for the 250 million smart meter rollout program. Smart meters record
meter readings every 15 mins and have the memory to keep the data for 45 days in the meter.
AMI requires two-way communications between the smart meter and the Discom’s computers in the control
room (or on the cloud). Various communication technologies, either individually or in combination, have
been used by utilities worldwide for AMI. Major utilities in North America, Australia, Japan, Nordic Europe,
South America, and South Korea have opted for the radio frequency mesh (RF Mesh) solution for their last
mile connectivity. Chinese4 and some European utilities have chosen power line communication (PLC)
technology, along with RF Mesh, for their last mile communication. Meanwhile, utilities in the UK5 and a few
other Scandinavian countries have adopted cellular technologies. Detailed features, architecture and
comparison of these different communication solutions are described in the next section 3.
Claim for IPR fees for using cellular technology for smart metering by cellular technology patent holders is a
new development that started in Europe. As the number of smart meters deployed on cellular technologies
scale up massively, such claims can arise in India as well. This issue is explained in APPENDIX-A
3 List of BIS approved smart meter manufacturing units in India is given in this link:
https://www.services.bis.gov.in/php/BIS_2.0/bisconnect/manufacturers/RGlyZWN0IENvbm5lY3RlZA==
4
Initially, smart metering with limited functionalities was rolled-out in China on PLC connectivity; later they experimented with RF
Mesh, Cellular and NBIoT technologies. The second-generation AMI which is about to begin in China is expected to deploy RF Mesh
for last mile connectivity
5
The UK launched the AMI rollout on cellular communications; but soon realized that cellular communication cannot reach meters
installed in the basement of buildings; hence they had to install RF based range extension systems. Out of 35 million smart meters
installed so far in the UK, about 40% of them are connected through RF communication
Head-End System (HES) and Meter Data Management (MDM) System are the most important software
solutions in an AMI system.
HES is a software that is responsible for fetching the meter data from the smart meters to the Discom’s
computers. Another important software for AMI system is the MDM which is installed in the Discom’s
computers where all meter data is collected and stored. In India, utilities record the meter readings every 15
minutes – 96 reads per day and this will be brought from millions of meters to the MDM by the HES. The
meter data organized in specified formats in the MDM helps to integrate it with the Discom’s billing system,
customer care system, geographical information system (GIS); and other IT applications. While there can be
several makes of meters, different communication technologies in different regions and several HES in a
smart metering solution of one utility, it is recommended to have only ONE MDM in a Discom which will
integrate all meter data with the billing systems and other Discom applications.
Having a single MDM for a Discom has several advantages over the use of multiple MDM systems. First, a
single MDM reduces integration costs by streamlining the process of data collection, analysis, and storage.
This eliminates the need for extensive customization and specialized staff training, resulting in significant
savings to the Discom. Another benefit of using a single MDM is the ability to standardize data reporting,
ensuring consistency and accuracy across the entire utility system. This not only simplifies data analysis, but
also improves system reliability and reduces the risk of errors, leading to an improved customer experience.
Additionally, the use of a single MDM system can alleviate the challenge of managing multiple competing
data systems. With a common data environment, Discoms can better coordinate their operational functions,
optimize resource management, and improve decision-making processes. In summary, the adoption of a
single MDM will help Discoms to streamline operations, reduce costs, and improve system reliability. By
consolidating data collection and analysis into one standardized platform, Discoms can successfully manage
the energy landscape and customer demands with greater efficiency and effectiveness.
Smart Meter Operations Centre (SMOC) is the control center with network monitoring system and advanced
analytical software solutions. The first-generation AMI projects did not have SMOC and as AMI data started
piling up, the utilities created SMOC with advanced analytical tools to handle the smart meter data. The time-
stamped meter-reads offer a goldmine of information about the power flows in the low voltage network
which helps to fine-grain the distribution grid including accurate demand forecasting that will reduce power
purchase cost and improve asset optimization. SMOC has proven to be beneficial in monitoring and
management of smart meter rollout as well. The standard bidding document (SBD) for the AMI program of
RDSS covers the functions of SMOC under Network Management Systems (NMS) and Network Operation &
Monitoring Centre (NOMC)6. Several functionalities and technical requirements of NMS and NOMC are
mentioned in these sections of the SBD. Detailed architecture for SMOC is not included in the SBD which is
left to the bidders to propose. For large scale AMI rollout, it is essential to have a well-designed SMOC that
can manage meter roll outs, data collection, data integration, data provisioning and data analytics. SMOC
typically provide/responsible for:
6
Section 6; Clause 2.2.2; and Clause 2.6 of the Model Standard Bidding Document for AMI
SMOC Analytics integrated with Customer Portal could provide real time visibility to customers on:
▪ Customer Billing and Energy Profile Information
▪ Prepayment Information
▪ Customer Centric Analytics
▪ Alerts on Account Information Updates
▪ Home Energy Profiles
▪ Customer Self Service
▪ Cost Saving Tips
▪ Customized Communication, Utility-branding, Customer Service and Feedback
Since most of the empaneled AMISPs are new to the AMI domain, it is recommended to include the detailed
architecture of state-of-the-art SMOC in the SBD.
In an ideal AMI system, there could be multiple makes of meters, different communication technologies,
different HES, but only ONE MDM. There are very few COTS7 MDMs in the world that are scalable to muti-
million meters. MDM software is expensive and its installation and integration with utility applications such
as billing system, customer care system, GIS; and outage management system is a very specialized job and it
takes minimum 6-9 months depending on the interfaces and the skill levels of the system integration team –
this is irrespective of the number of meters involved (whether 1 meter or 1 million meters).
There should be a standard middleware that should act as the integration platform through which different
Discom applications can call different data sets as and when required from the MDM. The selection and sizing
of the MDM and the middleware are very critical; and the hardware (on the cloud) sizing depends on the
software sizing. If the AMISP and their System Integrator (SI) get the sizing calculations (of software and
hardware) wrong initially, it will prove to be too expensive to correct them at a later stage. This is where prior
experience of handling humongous data generated by millions of meters and its integration with Discom
7 Commercially available Off the Shelf (COTS) refers to popular software products for which experts can be hired from the open
market. For proprietary software products one must always depend on the OEM for support. Most proprietary products may not
follow standard protocols and it may be difficult to integrate with other utility applications – all such integrations may be bespoke
developments which will be very difficult and expensive to maintain.
indows
ser Authen ca on Ac ve
illing ystems
M Directory
etwor
ostpaid
repaid
ntegra on ayer
rea ons Management
eather Data
App
eather Data
Applica on Monitoring etwor Monitoring
Meter ead re uest and commands, Meter and collector alarms vents and oll out schedule, nventory of nstalled Meters, or rders
meter data
ntegra on la orm epor ng
Meter procurement and inventory Meter details, reads, alarms events commands Analy cs
MDM
M M M M D
M
(SM1 – smart meters connected on RF mesh; SM2 – smart meters connected on cellular network; TSP NMS – network monitoring
system by telecom service providers for the meters connected on cellular communication; RF NMS – network monitoring system for
meters connected on RF mesh provided by the RF solution provider; DCU – data concentrator unit; GUI – graphical user interface;
WFM – workforce management system)
What is presented above is a Service Oriented Architecture (SOA) with micro-services which is the state-of-
the-art (SOTA) practice today. In the recent tenders from a few states, it is noticed that for each tender
package a separate MDM is provisioned which is not only expensive, but also prevents realization of several
benefits of smart metering. As mentioned already, with multiple MDMs (perhaps of different makes) in one
Discom, proper integration with billing system and energy accounting will be difficult. This approach is
erroneous and should be corrected immediately.
Public telecom networ or the utility’s fiber network (wherever available) is used for Wide Area Network
(WAN) solutions for AMI systems. The main challenge in a successful AMI system is the last mile connectivity
for which several solutions are available. The most successful AMI projects around the world have either
chosen RF Mesh technology or PLC technology for the last mile connectivity. In case of cellular technology,
the SIM card inside the meter is directly connected to the telecom network and there is no need for a
separate WAN network. The success of this depends on the data communication capabilities of the public
telecom network in a particular area. Different technologies available for last mile connectivity (NAN/FAN),
wide area network (WAN) and home area network (HAN8) are presented in the table below.
8Home Area Network (HAN) to connect the appliances inside home with smart meter was deployed during the first generation of
AMI. Now that smart apps are popular, the appliances can be connected to the broadband network at home and can be remotely
managed through apps on the smart phones; and their electricity consumption can be monitored through the customer portal of the
utility which has now become an integral part of the AMI system. Separate HAN is not being built these days.
▪ RF Mesh
▪ RF Mesh
▪ ZigBee ▪ Cellular
▪ 6LowPAN
▪ Wi-Fi ▪ Satellite
▪ ZigBee
▪ Millimeter Wave Bluetooth Low ▪ LPWAN
Wireless ▪ Wi-Fi
Energy (BLE) and BLE 5 ▪ Long Wave Radio
▪ Bluetooth
▪ Long Range Radio (LoRA) ▪ TV White Space
▪ Z-Wave
▪ Narrow Band IoT (NB IoT) -CAT ▪ Private Microwave Radio
▪ NFC
M1; LTE
Appendix-B presents a list of utilities who have implemented AMI around the globe and their chosen
communication technologies for their AMI systems.
Cellular Operators Association of India (COAI) has appealed to the Telecom Ministry regarding security risk
in deploying license-free spectrum for smart metering; and subsequently, Telecom Ministry has invited
a. Service Level Agreement (SLA): The AMI Service Providers (AMISPs) are bound to commit certain SLAs
(99%) for AMI system availability whereas none of the cellular service providers are ready to commit
guaranteed SLAs. Experiences with cellular communication for AMI in India indicate SLA below 95%
even in urban areas; which could be much lower in rural areas where data network is weak.
b. Technology Changes: Cellular operators upgrade their technologies frequently – 2G, 3G, 4G, 5G and
will move to 6G. Changing the network interface cards (NIC) and the SIMs in millions of meters is too
expensive and nearly impossible9. UK Government has recently sanctioned a £4 billion plan to replace
the 3G modems in smart meters with 4G, as the telecommunications companies will no longer provide
support for 3G. It is worth mentioning that, historically, the cellular telecom industry tends to undergo
upgrades every 5 years, leaving the previous infrastructure outdated which is closely aligned with the
handset replacement cycle.
c. Mis-match of Rollout Plans: Cellular operators primarily target urban areas for their new technology
rollout whereas electric utilities need to cover customers in urban and rural areas.
d. Cost: The initial installation cost of both RF Mesh (network interface cards + DCU/Gateway) and
Cellular (NIC + SIM) is nearly the same. But for cellular technology, there is a fee per meter per month
to be paid to the cellular operators which will be a huge burden that will eventually get passed on to
the common man (electricity consumer). Even at a modest fee of Rs 10 per meter per month, it works
out to about Rs 3000 crore per year to cellular operators for 250 million meters. The annual
maintenance fee for RF Mesh solution for AMI will be a small fraction (<10%) of that amount.
e. Reliability of Communication: Reliability of the cellular network in a given area depends on the user
density and when the utility wants to ping a particular meter, there is no guarantee that it will be
reachable at that moment as we often face with mobile calls not getting connected due to poor
bandwidth or network congestion. Last Gasps and First Breaths10 are to be logged by a smart meter
in case of power-off and power-on with in 20 seconds as per the SBD. When the Discom receives the
last gasp, they alert the maintenance teams on the power outage in a particular location and dispatch
the maintenance crew. This is one of the important benefits of AMI. With cellular connections, it is
9 In Uttar Pradesh, EESL deployed about 1.2 million smart meters with 3G during the period 2018 to 2020; and the telecom service
provider discontinued 3G service in UP. These meters are now operating on 2G as a fallback arrangement. New installations are on
4G; but future of 4G is uncertain as at some point, Telcos will stop 4G when 5G penetration achieve certain level of nation-wide
coverage
10 Last Gasp is the message of power outage communicated by a smart meter to the Discom; similarly, First Breath is the message
communicated to the Discom when power supply is resumed. Per SBD for RDSS, both Last Gasp and First Breath should be
communicated with in 20 seconds of the events. HES need to be programmed to act differently when one meter sends the last gasp
and when a large group of meters send the last gasp – if there is a power outage in a large community, thousands of meters will
send last gasp messages which the communication bandwidth may not be able to handle; and in such cases only select few meters
including the feeder/distribution transformer meters may be prioritized to transmit the last gasp. Such use cases should be
configured in the HES
RF Mesh solution providers should register with the Department of Telecommunications (DOT). The details
about the M2M Service Providers Guidelines published by DOT can be found on the website of DOT12. It is
recommended to include this in the SBD.
More recently, the telecom service providers have approached MOP (and REC which is the nodal agency for
the SBD) through Telecom Standards Development Society of India (TSDSI) to make changes in the SBD and
IS:16444 and IS:15959 so that cellular operators can meet the SLAs. This is another attempt of the telecom
operators to hijack the ongoing AMI rollout. They want to relax the standards so that their technologies (4G
and NBIoT) could meet the SLAs prescribed in the SBD. These attempts by the telecom operators must be
rejected.
As indicated in the beginning, AMI is not just to improve the metering and billing processes in a Discom. The
impact of AMI on the overall utility operations is much more as described in the table below. Several of the
benefits from AMI system depends on the reliability of the communication system deployed.
11 The 16 benefits to the Discoms and benefits to other stakeholders are presented in the Section- 4 of this paper
12The details about the M2M communication equipment registration can be accessed here:
https://dot.gov.in/sites/default/files/M2MSP%20Guidelines%20.pdf?download=1
There are always chances of human errors when meters are read
2 manually or even via automatic hand-held devices. In addition, the Medium Medium
process is time consuming. By delivering meter data automatically over
communication networks, AMI eliminates human error from the meter
reading process as well as make the entire process faster.
AMI can remotely detect meter tampering and enable real time energy
accounting. This reduces theft through by-passing the meter, thereby
6 High Medium
substantially reducing aggregate technical and commercial (AT&C)
losses. AMI will also streamline the billing, or meter-to-cash process
considerably by reducing the human errors in meter reading and billing
Better load research and demand forecasting from AMI data can
reduce power purchase cost
With meter data time stamped at 15-minute intervals, AMI enables near
real-time estimation of customer demand and understand customer’s
power consumption in granular detail. This improves DISCOM’s load
8 High High
forecasting and enhances the ability to procure the right volumes of
power. Utility can also implement time-of-use (ToU) tariffs for different
categories of customers and encourage load shifting with demand
response programs. These measures could reduce peak load and hence
reduce purchase of expensive power during the peak hours.
Asset optimization
15 Reduced load on call centres, customer care centres and billing centres Medium Low
Smart meters act as feedback points for understanding the behavioural
16 interpretations of energy demand as consumption which can be Low Low
modified
C. Benefits to Customers
1 Error-free bills and no need for visiting billing centers Medium Low
AMISPs should partner with best-in-class solution providers having standards-based security solutions. Key
standards include the NIST, NERC and ISO/IEC family of standards - ISO/IEC 27001, 27002, 27019 and 27035.
India has developed its own cyber security standards IS:1633513, focusing on operational technology (OT) in
the power sector. Smart meter standards IS:16444 and IS:15959, along with IEC:62056 series, define
communication protocols and associated security. IEC:62443 and IEC 62351 standards define the
compliances for cyber security for electrotechnical equipment and automation systems. In December 2021,
CEA issued Cyber Security Guidelines for Power Systems14 which should be followed by Discoms and the
AMISPs. Periodic security auditing, conformance testing of smart meters and devices, and cyber-security
lifecycle testing are essential to maintain compliance. Training the personnel associated with critical
infrastructure assets is important to enhance cyber security. Cyber-physical test beds need to be created to
test individual devices in integrated environments.
Some of the general security practices from the experiences of utilities who have implemented large scale
AMI in North America and Europe are summarized below for the considerations of Discoms and AMISPs.
a. Threats from Outside and Inside: Smart meters can be hacked by accessing onboard memory,
thereby reading diagnostic ports and other network interfaces. Besides, cyber criminals, employees
and vendors can unknowingly (or even knowingly) release sensitive information. To prevent such
attacks, utilities should impose intelligent controls on how employees, consumers and partners
access applications and data.
b. Security Principles: The global industry standard is the Confidentiality, Integrity and Availability - CIA
model of security. For AMI systems, Authentication must be added to this CIA model. Confidentiality
is to prevent sensitive data from reaching wrong people while ensuring that the right people still
have access. Integrity means data is consistent, accurate and trustworthy over the entire lifecycle;
and unauthorized people cannot alter data. This requires strong cryptographic mechanisms to ensure
the integrity of meter readings, command and control of the data. Availability of data and equipment
must be ensured by rigorous maintenance of hardware, prompt repairs; and upkeep of the software
free of corruption and conflicts. Firewalls and proxy servers could prevent downtime and mitigate
malicious actions such as denial of service (DoS) attacks. Authentication is to prevent unauthorized
access which happens often due to unmodified default access policies or lack of clearly defined access
policy documentation. Utilities must ensure that only authorized personnel can view information and
perform permitted actions. The HES, the field-tools and network devices must be deployed with
13IS:16335 is presently under revision and the updated version may be issued by end of 2023
14CEA’s Cyber Security Guidelines for Power Systems: https://cea.nic.in/wp-
content/uploads/notification/2021/10/Guidelines_on_Cyber_Security_in_Power_Sector_2021-2.pdf
c. People and Process: Insider attack is a key area of risk whether accidental or intentional. While an
outsider may be attempting to breach HES security which is being resisted by the system, it should
ensure that assigned employees are given legitimate access to the system. HES with role-based
access control (RBAC) may be deployed to provide capabilities to the Security Administrator to assign
appropriate permissions to each user of the system. HES could streamline user administration by
integrating with enterprise single sign-on solutions.
d. Data Protection: Meter data, customer billing information and other important data to be encrypted
end-to-end for maximum protection whether it is in a public or private cloud, on a device or in transit.
The end-to-end encryption help to combat advanced threats and maintaining regulatory compliance.
e. Advanced Security: Advanced security solutions should include signed and verified firmware,
disabled JTAG16-debug communications interface, encrypted flash memory, locked optical ports
(configurable), meter tamper detection, backhaul protection, certified root of trust; and other
physical and system level security features.
f. Key Management: The security solution must provide encryption key segmentation at individual and
group levels. Each end point (meters, DCU/gateway) is to generate its own AES 256-bit encryption
key to encrypt upstream and downstream messages sent to and from each end point. All the
individual keys of end points are vaulted in a Key Manager. HES can assign segment keys to a group
of end points. Device specific keys (protected through encryption) are stored securely during system
use and during rest. Device specific keys and network specific keys should follow configurable and
matured key rolling and lifecycle management processes.
g. Firmware Integrity: All firmware upgrades released are digitally signed using the utility’s private
key. Each end point within the network will validate the signature using the public key provided by
the HES. In case of signature mismatch, the end points will not upgrade the firmware.
h. Message Authentication: All commands may be signed with ECDSA17 standard using utility’s
private key. End points will execute signature validation before acting on any command, thus
providing a control mechanism to prevent rogue commands or man-in-the-middle vulnerability.
i. Third Party Penetration Testing: Utilities should engage certified third parties for penetration testing
to identify vulnerabilities and fix them periodically.
A comprehensive cyber security approach is crucial to safeguard the AMI system. For establishing secure and
15 Root of trust is ensured through Hardware Security Modules (HSM) where the Utility’s private key (encryption key) is vaulted
16
Joint Test Action Group (JTAG) is an industry standard for verifying and testing printed circuit boards after manufacture. It gives a
pins-out view from one IC pad to another so that faults could be discovered. JTAG became the IEEE 1149.1 standard in 1990.
17
Elliptic Curve Digital Signature Algorithm (ECDSA) is a variant of the Digital Signature Algorithm (DSA) which uses Elliptic-Curve
Cryptography (ECC). ECC uses much smaller public keys compared to the other popular encryption methodology called RSA. For AES
128 encryption, the ECC key is 256 bits whereas RSA key is 3072 bits; and for AES 256 level encryption, the ECC key is 512 bits (64
bytes of 8 bits each) while RSA key is 15360 bits. ECC needs much lower processing power and gives faster SSL handshaking and
consequently faster web page loading.
a) All the hardware, operating systems and application software should be hardened
b) Application, scanning and hardware scanning tools should be provided to identify vulnerability and
security threats
c) Data should be encrypted at system/device/technology level
d) Network zoning should be implemented as per the proposed architecture (or other methods of
network architecture without compromising the security of the system)
e) Internal users should be allowed to access all adjacent zones - they will not have access to remote
network zone
f) While procuring cyber security items testing must be done and the system must be secure by design
g) Residual information risk should be calculated by AMISP and same should be submitted to the
Discom for approval
h) All default user ID and passwords should be changed
i) All log in/out and cable plugs in/ out should also be logged in Central Syslog Server
j) Penetration and vulnerability assessment test by CERT-IN certified auditors during SAT and operation
and maintenance period
k) Auditing by third party during SAT and annually during operations and maintenance period should
be in the scope of AMISP18
l) As the computer system in NOMC (SMOC) has access to external environment, the AMISP should
document and implement Cyber Security Policy/Plan in association with the Discom to secure the
system
m) Discoms and AMISPs to follow the latest Cyber Security Guidelines issued by CERT-In
(http://www.cert-in.org.in/); and the provisions under “Testing of all equipment, components, and
parts imported for use in the Power Supply System and Network in the country to check for any
kind of embedded malware /trojans/ cyber threat and for adherence to Indian Standards –
Regarding” vide rder o. o.9 6 0 6-Trans-Part(2) issued by MOP on 18 November 2020 and
amended from time to time or any other competent authority
n) AMISP should adhere with the appropriate security algorithm for encryption and decryption as per
established cyber security guidelines. For smooth functioning of the entire system, it is essential that
the AMISP shall provide in the form of a document enough details of such algorithm including the
mechanism of security key generation to the Discoms. In case of proprietary or secret mechanism,
the same shall be kept in a secured escrow account.
Similarly, Section 2.7.8 of the SBD prescribes the measures to be adopted for data privacy and data security.
AMISPs should ensure that the system is compliant with the applicable provisions of the “Reasonable security
practices and procedures and sensitive personal data or information Rules, 2011 (IT Act ” as well as should
be committed to work with Discoms for compliance to personal data protection requirements. The Discom
should be the sole custodian of the smart meter data19. The AMISP and its contracted vendors will have
18 We hope all Discoms have incorporated these clauses in their RFPs and contracts
19 By law many countries have established that the consumer is the sole owner of the smart meter data; and Discom is the custodian
SBD Section 2.2.3 describes the measures for Network Protection and Security; and section 7.4.1 mandates
that the AMISP should be responsible for monitoring of the system from a cyber-security perspective. The
logs of the system shall be analyzed for exceptions and the possible incident of intrusion/trespass should be
informed to the Discom and analyzed to discover the root cause. The monitoring should encompass all cyber
security devices installed in the cloud data center as well as at the NOMC (SMOC) such as firewalls, all types
of intrusion prevention systems, routers etc. The cyber security system should also be subjected to Annual
Security Audit from CERT-In approved auditors at the cost of the AMISP during the contract period. AMISP
should share with Discom such audit reports and implement the recommendations/remedial actions
suggested by the Auditor. Again, we are afraid whether all the AMISPs who are bidding in different Discoms
have any idea about what it takes to comply with the above provisions in the SBD related to cyber security.
Telecom Engineering Center (TEC) under the Department of Telecoms (DOT) has published the Technical
Report on Security by Design for IoT Devices Manufacturers (TEC 31328: 2023)21. AMISPs and their project
partners may be mandated to study and comply with these procedures as well. REC may consider adding this
provision in the SBD.
As part of R-APDRP, 14 Discoms in India were allotted smart grid pilot projects in 2013. Out which only 11
projects have been completed and all these projects had smart metering ranging from 1200 to 30000
customers. Most of these projects took 4-5 years to implement. Having observed the trials and tribulations
of Discoms with these first set of smart metering projects, ISGF was convinced that the state government
owned Discoms will not be able to procure right AMI systems; and even if they install the right systems, they
will not be able to maintain it for ten years. ISGF published a White Paper in 2016 (which was re-issued in
March 2017 as a joint paper by ISGF and BNEF) that articulated the idea of engaging a Metering Services
Agency (MSA) who will install the AMI system and maintain it for 10 years for a monthly fee per meter. This
is the same business model which is adopted for the 250 million smart metering projects under RDSS. Only
difference is that under RDSS, GOI is giving 15% of the project cost as a grant; and the rest is paid in monthly
installments over 93 months. In our original paper we estimated the cost of a single-phase smart meter at
INR 2250 and the MSA service fee at INR 69 per meter per month for ten years for a project with one million
(or more) meters, which was about one US dollar per meter per month in those days. Today under RDSS, the
average price being quoted by AMISPs is about INR 80 per meter per month for 93 months (which is about
one US dollar).
20 AMISP should commit to ensuring that the data is kept safe by them and their sub-contractors/project partners and not used for
any other purpose
21 https://tec.gov.in/pdf/M2M/Security%20by%20Design%20for%20IoT%20Device%20Manufacturers.pdf
In our opinion this is neither practical nor logical to have 250 million smart meters operating in the
prepayment mode. In most Discoms, the high-value customers contribute 70-80% of the revenue who may
be only 15-20% in numbers; and majority of them pay several million rupees per month22. Moving them to
prepayment mode will have commercial and technical challenges. Regular customers who consume above
500 kWh per month (or an appropriate limit set by the state regulator) may be allowed to opt for either post-
paid or pre-paid modes. Those opting for pre-paid in this category may be offered a small rebate to motivate
them. All customers with less than 500 kWh monthly consumption and government offices may be brought
under mandatory prepayment mode. Even their numbers may be in millions in most Discoms.
The smart meters can be configured in either prepaid or post-paid mode. But the trouble with prepayment
operation of smart meters is that most of the HES are not designed to respond to a large number of
disconnect/reconnect requests in less than ten minutes as prescribed in the SBD. If thousands of customers
recharge their meters online, immediately a reconnect order will be generated by the system; but it will get
into a que in the HES – particularly when HES has already issued a command to download the interval read
of millions of meters. In this scenario, the recharge may be updated and electricity supply resumed after few
hours. This can be fixed to certain extent by modifying the HES provided the communication system is
reliable; but most of them would still find it difficult to meet the SLA of ten minutes to reconnect supply after
recharge.
Each Discom in consultation with their state government and respective electricity regulatory commission
may decide what all categories of customers in which all regions should be brought under prepayment mode.
MOP may allow the states to take this decision as appropriate. Afterall, the net-grant from GOI for AMI under
RDSS is only 6% of the project cost23.
It is understood that when the RDSS program was launched, the project cost was calculated at INR 6000 per
meter and accordingly the 15% GOI grant was capped at INR 900 per meter. There is confusion about this
number while one argument is that this amount of INR 6000 was the capex cost under the EPC model of
project implementation; while the other argument is that it was the life-cycle cost of AMI implementation
that EESL offered in UP and Haryana in 2017 which was for 6 years (72 monthly installments versus the 93
monthly installments in RDSS). In our view, the life-cycle cost needs to be revised to the range of INR 9000 to
12000 per meter depending on the geographical challenges and the total number of meters involved. The
prices could be higher for very low volume contract packages (below 200,000 meters) as well as for very high-
volume contract packages (above 5 million meters)24.
22 Large C&I customers have HT or LT-CT meters which have no built-in switch for disconnect-reconnect operations; and hence
cannot be moved to prepayment mode
23
GOI is offering 15% grant under RDSS for smart metering; but collects 18% GST on the project cost including the monthly
installments. Out of this 50% is passed on to the state governments; hence the net-grant from GOI for smart metering is only 6% of
the project cost
24
For a project with 5 million meters, even 1% of the meters that cannot be read in a month will be about 50,000 and manually
This is another big question haunting the decision makers in the states. In some states the Discoms have
approached their regulators for a pass-through in the tariff for the monthly installments. Their argument is
that in the initial years Discoms may not be able to bear the additional burden of the monthly installments
to be paid to the AMISPs as it would take few years to realize the full benefits of the AMI system; and such
efficiency gains when realized will be passed on to the customers in tariff relief. If adopted, this approach will
place a huge burden on low-income communities whose electricity bills are in the range of INR 200-300 per
month which will go up by another INR 80-100 per month. The promised tariff reductions after a few years
will not motivate them to get their buy-in for the AMI program. In some states where the existing metering
and billing systems are in poor condition, the benefits of AMI can be realized right from the very beginning
through increased revenue per month which itself will take care of the monthly installments25.
We suggest to offset this monthly fee by dividing it in to 3 buckets – one part may be added to the meter
rent that all Discoms levy in the electricity bills; another part may be borne by the Discoms and the third part
may be funded through a low-interest loan from REC/PFC which may be paid back from the efficiency gains.
The percentage of each of these 3 parts may be decided by each Discom in consultation with their
governments and regulators.
Ideally, AMI rollout should start with one city/division in a Discom which has about 1 million meters. This first
contract package should have MDM and system integration components; and the system integrator (SI)
should successfully integrate the MDM with the billing system and other applications and test it. Once the
backend systems are stabilized and the first batch of meters (>100,000) can be read remotely and the
monthly bills can be generated (without human interventions), the Discom should engage multiple agencies
to rollout the smart meters in other cities/divisions. Those new contract packages should have smart meters,
communication, and HES (no MDM). The responsibility for integration of the new HES with the MDM should
be under the scope of the SI of the first contract package.
We recommend a three-phase rollout – first in the pilot city with about 1 million meters, next in all other
urban and semi-urban areas; and lastly extend to the rural areas. This approach gives time to the Discoms to
extend their billing system to non-RAPRDRP towns and rural areas that are still having multiple billing
systems. AMI implementation may not be feasible in several hamlets and habitations in the hill areas and
tribal communities in the forests. Discoms may be allowed to decide which are the pockets/communities
where AMI is not feasible. In such cases, Discoms may install smart meters for feeders (wherever feasible)
and monitor the community’s consumption to prevent misuse.
Some of the Discoms are making changes in qualifying requirements, SLAs and other important parameters
specified in the SBD. These are primarily vendor driven. REC and PFC (MOP) should stop such changes in
reading them in a diverse geography will be very expensive; and these unreadable meters are not the same every month
25
In one of the states, it was estimated that the average revenue growth after AMI implementation was above INR 200 per month
per customer while the monthly installments to the AMISP was well below INR 100 per meter per month
In our observation, the preparedness of Discoms for large scale AMI rollout is still lacking in almost all states.
Many states do not have a single Discom-wide billing system which is a prerequisite for successful AMI
deployment. Home grown billing systems that exist in several states may not be easy to integrate with
standard MDMs. Too much of customization of the MDM will diminish its true potential; and prove to be too
expensive to maintain in the long run. Discoms do not regularly update their GIS maps with customer
indexing. Also changes in the field are not communicated to GIS team hence even after the consumer
indexing is done the data mismatch appears. Adequate manpower and other resource constraints continue
to haunt the Discoms. These issues need to be addressed on priority to reap the benefits of the AMI system.
What should be the ideal life of smart meters? As per the BIS certificates issued to the meter manufacturers
it is mentioned that 5½ years of warranty from the date of delivery or 5 years of warranty after the meter is
installed. Hence, typically meter OEMs in India have been giving 5 years warranty. Now for the RDSS projects,
most of the OEMs are offering up to 10 years warranty.
Utilities in USA and Europe mandates minimum 12-15 years life for meters. This is one way of reducing the
overall cost of smart metering. One issue that could hamper long life of smart meters in India is the battery
life26 in high ambient temperatures.
We recommend that BIS amend the certification with minimum 10 years warranty; and mandate highly
accelerated life test (HALT) for meter-life expectancy testing in India.
In general, very few agencies in the country understand the complexity in installing multi-million-meter AMI
systems and maintaining them for nearly 8 years (93 months as per SBD). AMISPs with no prior experience
of smart metering are signing up for implementing 5-6 million smart meters in 28 months! Most of the
Discoms and AMISPs do not seem to be taking any serious efforts to train their engineers in AMI.
This reminds us of the R-APDRP Part-A projects that were awarded by Discoms during 2008-2012. There was
a set of System Integrators (SI) who were empaneled by PFC – mostly large IT companies, both domestic and
foreign. The project implementation time specified was 18 months. R-APDRP Part-A scope included indexing
of Discom’s assets and consumers in 4 towns on the maps which required thousands of trained
technicians who could handle DGPS equipment (which was the only way to do GIS mapping those days); and
there were not even few hundred trained technicians in India at that point in time. Hence, all foreign IT
companies stayed away from bidding for R-APDRP projects. Large Indian IT majors competed aggressively
and signed up to execute projects within 18 months at prices way lower than the amounts budgeted by MOP.
However, all R-APDRP projects took 5-7 years to complete; and all the Indian IT companies who executed
those projects incurred huge financial losses. None of them are participating in the ongoing AMI project
tenders.
26All smart meters have lithium-ion batteries that lasts typically 10-12 years in moderate temperatures. At >45° centigrade
temperatures the long-life expectancy of these batteries is doubtful.
MOP may review the situation and take appropriate measures to ensure that personnel engaged in AMI
projects from both Discoms and the industry are given proper training. The Part-C of R-APDRP had over INR
2 billion for training and capacity building; but actual spend was a minuscule portion of that. PFC engaged
agencies with no prior experience to develop training modules at very low cost. The results and experience
of R-APDRP are evident. We strongly advocate for spending minimum 5% of the project cost of the RDSS
program in training and capacity building for Discom personnel so that 95% of the investment is well spent
and the intended benefits are realized. It is high time for GOI to realize that in areas like training and capacity
building in emerging technologies where best-in-class agencies must be engaged, procurements cannot be
done on the regular L1 bid route.
g. Customer Engagement
For successful AMI rollout and customer’s participation in leveraging the full benefits of the AMI systems, it
is essential to have customer engagement in the program right from the beginning. In many countries
customer groups opposed smart metering. In USA, 15 states had to include Opt-Out option in their AMI
programs because of customer objections; and in many countries AMI rollouts were suspended mid-way and
engaged in long consultations with customer groups for their buy-in. We do not see customer engagement
activities in the ongoing AMI rollout in any of the states in India so far.
7 AMI 2.0
As stated in the beginning, Indian Discoms have the great opportunity to leapfrog to AMI 2.0 as we have done
with our mobile telephony two decades ago. The new features of AMI 2.0 which was not there in the first set
of smart metering projects are additional functionalities that can be realized at marginal cost as explained
below.
a. Advanced Analytics: The time stamped electricity consumption data captured from smart meters
can be analyzed with the help of Artificial Intelligent (AI) and Machine Learning (ML) tools to
understand the power flows in real time and identify overloaded/stressed assets; locate which
transformers to be replaced with higher capacity ones; which transformers have phase imbalance
issues that must be corrected27; detect meter tampers and irregular usage patterns; detect
27
In an AC distribution network, typically loads are segregated amongst the 3 phases based on contracted load; but in real-life, load
on one or two phases may be much higher than what was allocated; this causes overloading of the distribution transformer and it
could even lead to transformer burn-outs
In conclusion, we suggest Discoms and AMISPs should brainstorm and design their AMI systems such a way
that additional functionalities could be built in to the system at marginal cost which will be additional revenue
streams for both Discoms and AMISPs which will eventually reduce the burden on the electricity rate payers.
MOP or CEA may like to constitute a committee of technical experts and select utilities to review and suggest
measures to undertake a course correction in the ongoing AMI program.
As part of the 250 million smart meter rollout, as many existing old meters (non-smart) will be taken-out
from the customer premises. The model SBD mentions that the old meters should be deposited with the
Discoms. The evolving practice on electronic and other hazardous materials recycling is through Extended
Producer Liability programs in which the producer is liable to take back the product at end of life and
recycle/reprocess it in a scientific manner without emitting or dumping of any hazardous materials in the
environment. The Battery Waste Management Rules 2022 issued by the Ministry of Environment, Forest and
Climate Change (MoEFCC) for disposal of lithium-ion batteries clearly define the responsibilities of the
producer, consumer, public waste management authorities and the recycler.
Majority of the old meters (more than 200 million) are electronic meters which need to be recycled or treated
like other electronic waste. Among the electronic meters installed in India there are two generations – old
meters produced before 2004 which have lead (Pb) in their printed circuit boards (PCB). The later versions
are lead-free. The meters with PCBs having lead is very hazardous to the environment. Both these categories
of meters need to be segregated and send for recycling/disposal. All the electronic meters have batteries
(mostly lithium-ion batteries) which also need to be taken out and recycled separately. Discoms do not have
the bandwidth to do such segregation and disposal in a scientific manner; hence this responsibility for safe
disposal of old meters may be assigned to the AMISPs or the policy makers should help creation of a recycling
industry to handle this huge task.
As millions of smart meters connected on cellular telephone networks are deployed in India, there is a risk of
hidden cost towards IPR fees that may be claimed by the technology companies who own these IPRs in 3G,
4G, 5G, and NB IoT technologies. In the journey towards achieving smarter energy grids, it is undeniable that
cellular technologies are playing a pivotal role. Numerous Indian companies are actively developing and
deploying smart meters that rely on cellular networks to transmit data. But as the number of meters
connected on cellular networks increases, the chances of such claims by patent holders are more likely. The
complex landscape of Cellular Standards Licensing is explained here.
The process of licensing cellular standards and the associated IPR fees is very complex, but it is crucial for
businesses to understand this landscape. IP Europe (www.ipeurope.org) is combining information on
intellectual property and principles, how the licensing is managed globally, and how they are applicable to
India as well. IP Europe is supporting the CEN-CENELEC Workshop Agreement (CWA) “ rinciples and
uidance for the icensing tandard ssential atents in 5 and the nternet of hings o ”. This document
(CWA17431) describes the licensing principles of the major Standard Essential Patent (SEP) owners and how
they are exercising their IPR portfolio. The general principle is Fair, Reasonable and Non-Discriminatory
(FRAND) terms, meaning the license must be granted without discrimination towards any party. It also
defines the value of the patent which is based on its value to the (end) users. That means the cost of license
varies depending on the final use case and the value of the use case. Examples of different use cases could
vary from automotive, healthcare, energy, and the financial sectors. Following the FRAND licensing principles,
similar practices are valid for the smart metering use case which apply technologies to enable the seamless
functioning of cellular networks.
Already, we are witnessing examples of cellular IPR holders like Nokia engaging with IoT product makers who
have integrated cellular technologies into their products and therefore are entitled to collect IPR fees. Nokia
has won a case against Daimler for using cellular connectivity in automobiles which is now binding globally.
We also hear that some of the Standard Essential Patent (SEP) holders are already demanding IPR fees from
smart metering companies/utilities in Europe for deployment of smart meters with cellular connectivity. The
expectation is that similar scenarios will soon unfold in India. With the surging adoption of cellular-based
smart meters, the question of determining the value of cellular communication patents in the context of
smart metering becomes increasingly pertinent. Nokia, Ericsson, Huawei and Qualcomm are important SEP
holders in cellular, all making multi-billion-dollar annual revenue from IPR licensing.
It is important to recognize that the value of cellular communication patents in the realm of smart metering
cannot be understated. The CWA17431 document recommends engaging with the SEP holders early enough
to collect information on their demands of licensing.
As already mentioned, Nokia, Ericsson, Huawei and Qualcomm hold majority of the patents. According to the
ETSI IPR Online Database and ABI Research analysis, roughly 172,000 3GPP/5G declarations on essentiality
have been made as of 2022, covering 51,000 unique patent families. China leads the pack for the number of
5G declared patent families. In every cellular network, multiple patents of different companies are deployed;
and each one of them can claim their IPR fee separately. So far there are no inter-company agreements for
collectively claiming IPR fee and sharing amongst these companies.
As Indian companies forge ahead with their smart meter deployment strategies, they must factor in the
potential costs associated with cellular IPR fees. Integrating these costs into business plans is not only a
prudent step but also a necessary one. By being prepared to address these fees, companies can ensure the
viability and sustainability of their smart meter initiatives and minimize potential disruptions and financial
challenges down the line.
Conclusion
The smart metering deployments are propelling India towards a greener and more intelligent energy future.
However, as the landscape evolves, it is essential for all stakeholders to recognize the significance of IPR fees
when choosing to use these technologies. By understanding the licensing processes, anticipating potential
encounters with patent holders, and factoring in IPR costs during the planning phase, companies can position
themselves for success in the rapidly evolving world of smart metering. IPR licensing fee should be considered
in overall project costing and ROI calculations by Discoms and AMISPs if they are deploying smart meters
connected on cellular networks. Ideally, AMISPs (or the communication services providers) should indemnify
the Discoms from potential claims of IPR fees by SEP Holders. In this dynamic era, staying informed,
adaptable, and strategic will be key to harnessing the full potential of smart metering while navigating the
complexities of intellectual property rights and associated fees.
APPENDIX – B: List of Utilities who have Implemented AMI and the Communication
Solutions Adopted
This list is compiled from publicly available data and may not be exhaustive and very accurate. There could
be several more utilities who would have undertaken AMI in the recent past. Most of the data in this list is
updated up to 2021.
Communication
Number of Year of Project
Sl No Utility Name Technology
Meters Execution
Deployed
A. USA
1 AEP (APCO) 1,100,000 2017 RF
2 AEP (PSO) 530,000 2014 RF
3 AEP (SWEPCO) 560,000 2020 RF
4 AEP Texas 1,041,000 2017 RF
5 Arizona Public Service 1,200,000 2012-2014 RF
185 Yazoo Valley Electric Power Assn. 10,561 2020 RF and PLC
Total 103,827,999
B. Canada
1 ATCO Electric 100,000 2020 RF
2 Alectra 1,000,000 2023-2026 RF
Total 1,900,000
AE. Chile
1 CGE Distribución 1,400,000 2014-2018 PLC
2 Enel Distribución Chile 2,400,000 2017 PLC
3 Luz del Sur 500,000 2014-2019 PLC
4 Chilquinta Energía 600,000 2019 Cellular
Total 4,900,000
AF. Norway
2,220,000 2018 RF
1 Norway
670,000 2018 RF
2 Elvia 950,000 2016 RF
3 Glitre Energi Nett 100,000 2018 RF
4 Lede 220,000 2014 RF
5 Lnett 160,000 2014 RF
6 Lyse 140,000 2014 RF
7 Midtkraft Nett 14,000 2016 RF
8 Nettselskapet AS 37,000 2019 RF
9 Trondheim Electric 50,000 2021 PLC
10 Norway 330,000 2020 Cellular
Total 4,891,000
AG. Croatia
Total 400,000
AH. Cyprus
1 Cyprus Power 70,000 2019 RF
2 Nicosia Energy 60,000 2020 RF
3 Mediterranean Utilities 55,000 2018 PLC
4 Limassol Electric 65,000 2021 RF
5 Paphos Power 50,000 2017 Cellular
Total 300,000
AI. Czech Rep
1 Czech Power 120,000 2018 RF
2 Prague Energy 90,000 2019 RF
3 Bohemian Utilities 80,000 2017 PLC
4 Moravia Electric 100,000 2020 Cellular
5 Vltava Energy 70,000 2021 RF
Total 460,000
AJ. Denmark
1 Danmark Power 150,000 2019 RF
2 Copenhagen Energy 120,000 2020 RF
3 Nordic Utilities 100,000 2018 PLC
4 Aarhus Electric 80,000 2021 RF
5 Zealand Power 90,000 2017 Cellular
Total 540,000
AK. Hungary
1 Magyar Power 120,000 2018 RF
2 Budapest Energy 100,000 2019 RF
3 Danube Utilities 90,000 2017 PLC
4 Transdanubia Power 80,000 2021 RF
5 Pannon Electric 110,000 2020 Cellular
Total 500,000