Smart metering experiences – the foundation for further smart grid innovation


Scandinavia led the way for Advanced Metering Management (AMM) implementations in Europe in 2003 when Sweden announced its decision to require monthly meter readings by 2009. It spurred activities in all the other Scandinavian countries, resulting in deployments in utilities like Vattenfall, Skagerak and Fortum, to name but a few. As Sweden approaches full deployment, some interesting patterns and observations are becoming apparent which may hold a key for future developments in the move towards smart grids.

In a meeting with many of these utilities approaching the final phase of their large-scale smart meter rollout, we observed the following: Half of the utilities stated that when the rollout was finished, the job was done, and they would start automated meter data collection. The other half of the utilities, however, pronounced that when the rollout was finished their real job was about to start – and they looked forward to it! What did they look forward to – and why? This observation and the reply to this basic question deserve some comments and explanation.

In the discussions between the utilities we observed both commonalities and differences. During the conversion from being a utility with Advanced Meter Reading (AMR) for large Commercial and Industrial (C&I) customers to becoming a utility with all customers linked to an Advanced Metering Infrastructure (AMI), all of them had of course focused on planning, purchase processes, meter rollout, system integration and meter data collection. How these focus areas were managed can be categorised depending on their number of customers.The table below gives a summary of the observations. The medium size has 50,000 – 200,000 customers while the few large ones had more than 200,000 customers.


Typically this statement is observed coming from utilities where the functionality is primarily regulatory driven and where the department of meter operation has the role of managing the whole process from planning to data collection. Utilities in this class very often choose one major meter vendor, in which case they seldom install a stand-alone meter data management system (MDMS). Meter deployment is often done by a combination of employees and local installation companies. When deployment is finished, the meter department has the responsibility for data collection and the export of data to the billing system. If they do not have an MDMS, the validation is executed either in meter vendor data collection software or the billing system. System integration typically is by file transfer, but they are considering moving toward real-time information exchange using web services – without any integration middleware. For data collection, the topology of the supply area is the basic premise for choosing the type of communication technology. Except for a few companies using only RF, the typical situation is a mix of PLC, GPRS and RF to some extent. This observation was made primarily among utilities with fewer than 50,000 customers, the major class of utilities in Scandinavia.


The people with this energetic perspective and desire to develop beyond smart metering typically work for medium and large utilities. As can be read from the table, they have a significantly different approach to both planning, purchasing, rollout and system integration than small utilities. Smart metering is a strategy involving the whole company. The legal framework, in combination with their strategic actions, opens up for smart metering expansions towards smart grid and other new services. In this process they focus on system integration between business systems. When it comes to meter data collection the large utilities have achieved valuable experience regarding capacities and technologies. With the availability of new system integration tools, utilities have created new ways to work smarter and improve operational efficiency based, in part, on the valuable information gained from large-scale meter data collection.

“…the role and function of the different software systems involved must be described in detail.”


Smart grid features – an important part of a smart metering project The movement from AMR of C/I customers to deployed smart metering followed by an innovative phase towards smart grid features is observed among several of our utility clients. The step from smart metering to smart grid is typically a result of newly available technical features in the meter combined with company strategies for adding value to the smart metering investment. In addition, it is crucial that the system integration between the business systems has the ability to serve all those new smart grid features. The different steps of this growth and innovation strategy are described in figure 1.

The realisation of smart grid features includes a strategy of meter deployment in transformers and grid exchange points in addition to the end-use customers. The first smart grid features demanded are typically management of alarms and events in addition to the possibility of serving the grid operation centre with updated metered information from the whole grid – including balance reports and results prepared for the customer call centre and third-party business systems.

System integration –and an exact description of the roles of business systems.

A lesson learned from efficient operation of large-scale smart metering projects is that the role and function of the different software systems involved must be described in detail. It is also observed that this process must be followed by a definition of master data for each system. This definition typically causes some internal noise and intensive discussions in the utilities. Should the billing system still be the master for information about meter and communication devices – or should it be the MDMS or NIS/GIS system? Typically the conclusion is that MDMS is the best system to manage all information about the meter devices – while the billing system is the best master for all customer-related data. Figure 2 shows how the basic software systems in a large-scale smart metering project are given responsibility as a total solution integrated through a bus solution. The CRM system is not included in the figure.

The solution is established – let’s start operation!

Operation of large-scale smart metering projects is still in a phase of evolution needing evaluation and fine-tuning before efficiencies can be realised and new business opportunities can be created. Some operational experiences from the Nordic market are available and have provided valuable know-how and conclusions. The first question in operations is typically about the necessity for time resolution and frequency of data collection. Daily collection and the possibility to configure the meter for hourly resolution fulfil the basic needs in the short term. However, in the last year we observed that the grid companies ask for different kinds of event logging (outage, voltage, theft detection etc) and this has resulted in the need for event management functionality in the MDMS. A basic requirement for all data collection systems is the ability to execute on demand meter readings. In the Nordic market there has never been a discussion regarding the need for 15-minute interval meter reading of consumer metering. However, a 15-minute interval is of relevance when it comes to an imbalance in market/ancillary services only – and does not influence smart metering requirements.


Figure 1.

 As a software vendor of head-end and meter data management systems, Powel, together with its utility customers, has gained valuable experience in several of the focus areas listed in the table. Some of these experiences are general and are listed here:


    Figure 2.
  • When a new meter is installed, the customer is often sceptical and asks, “What’s in it for me?” The result is an increased number of contacts with the call centre in the first phase of the project. However, after some months the number of contacts is reduced; and most utilities report that one year after deployment it is 20-30% lower than before deployment.
  • By using always off technology, data collection from 465.000 GPRS meters with daily resolution takes 7 hours with two communication servers. In addition, we use one validation server and one database server. Data collection is performed every day.
  • Data collection from 150.000 MESH meters with daily resolution is done daily. For this technology, the whole day is used for collection. All data is temporary stored in tables before it is validated once a day and stored in a database. In this project two communication servers and one validation server are used.
  • Data collection from GPRS meters typically has a hit rate of 99.7 % for the first reading. Only 0.3% of the meter users need to be sent smses to ask for a re-reading.
  • Some utilities have installed meters that send messages about events and alarms. To manage this information, it is crucial that there are functions for filtering and analysing the information before sending it as an alarm requiring action. An example is that Powel’s head-end system for a customer with 100.000 meters received 10 million alarms during one week. Reasons can be: a faulty setting of threshold values; the quality of supply is lower than expected; inaccurate meters; or the log on/off to the communication network disturbed the meter.

To be able to do the data collection efficiently, the communication network must be operating well and safely. We learned that this may be a challenge due to the fact that utilities are not traditionally communication network operators. They often underestimate the cost of operating PLC and radio networks. Regarding GPRS, they buy the service from a telecom operator.


Utilities act according to their size. Deployment of smart meters is experienced to be in line with their installation schedule. Software systems for data collection and meter data management are a part of the total solution that includes several business systems. The responsibility of operating PLC and RF networks is observed too often to be a new challenge for utilities. Data collection, even for largescale projects, is not regarded as a problem. In companies where strategy is the main driver for the introduction of smart metering solutions, the gap between metering and smart grid operation is quickly bridged! In utilities where the roles of the different business systems are well defined and the system integration is working or planned for the required performance of smart grid features, the operation of fully deployed smart metering projects is rapid, together with smart grid operation. The one critical factor of success that should be emphasised is that smart metering and smart grid utility staff cooperate very well.