multiFLEX: Flexible Multi-utility, Multi-service Smart Metering
Architecture for Energy Vectors with Active Prosumers
Edoardo Patti
1
, Enrico Pons
2
, Dario Martellacci
3
, Federico Boni Castagnetti
3
, Andrea Acquaviva
1
and Enrico Macii
1
1
Dept. of Control and Computer Engineering, Politecnico di Torino, Torino, Italy
2
Dept. of Energy, Politecnico di Torino, Torino, Italy
3
IREN Energia, Torino, Italy
Keywords:
Distributed Software Infrastructure, Smart Metering, Smart Grid, Demand Response, Distributed Systems,
Distribution Network.
Abstract:
In order to move forward the vision of Smart Grid, a flexible multi-utility and multi-service metering archi-
tecture is needed to allow innovative services and utilities for the different actors playing in this scenario. To
achieve this, different meters (e.g. electric, water, heating and gas meters) must be integrated into a distributed
architecture in order to gather and analyse heterogeneous data. Hence, such architecture provides in real-time
a complete overview of the energy consumption and production in the grid from different prospectives. From
customer viewpoint, this information can be used to provide user awareness and suggest green behaviours,
thus reducing energy waste. From energy operator or utility provider viewpoint, for instance such analysis
can: i) improve the demand response for optimizing the energy management during peak periods; ii) profile
consumer energy behaviours for predicting the short term energy demand; iii) improve energy and market
efficiency. In this paper, we discuss the characteristics of this infrastructure and its expected impacts on utility
providers, energy operators and customers.
1 INTRODUCTION
The electricity market was introduced in the Euro-
pean countries following the Directive 96/92/EC of
the European Parliament and of the Council concern-
ing ”common rules for the internal market in electric-
ity” (European Parliament, 1999). Up to now the mar-
ket is working properly for big producers, retailers
and users, while the small consumers and prosumers
cannot access directly the market and cannot be influ-
enced by price signals.
Distributed generation from renewable and
non-programmable energy sources is becoming
widespread. This requires a more flexible manage-
ment of distribution grids, also involving energy
storages, both at the prosumers and on the network.
For these reasons, a smarter” grid is needed by
Transmission and Distribution System Operators
(TSOs and DSOs) together with retailers and market
operators in order to face in their activities.
A first step in this direction is the development
of a flexible smart metering architecture for multi-
ple energy vectors (multiFLEX). In some countries,
smart meters are already deployed at user level. Such
smart meters are nearly devoted to billing improve-
ments. However, a new metering systems is needed
to go much further by providing their contribution to
various objectives such as: i) end-user affordability
of electricity; ii) energy and market efficiency im-
provement; iii) CO
2
emissions and pollutants reduc-
tion. Hence, such flexible smart metering architecture
must be able to: i) integrate the already available com-
ponentsand devices that can be implementedin a plug
and play way; ii) combine and correlate information
from meters of different services such as electricity,
water, gas and district heating; iii) provide advanced
services to users, DSOs and other utilities; iv) en-
hance the retail market. Following this view, the mul-
tiFLEX infrastructure integrates different information
from heterogeneous data-sources to promote innova-
tive services related to electricity, water, gas and dis-
trict heating.
The rest of this paper is organized as follows. Sec-
tion 2 reviews the relevant state of the art on dis-
288
Patti E., Pons E., Martellacci D., Boni Castagnetti F., Acquaviva A. and Macii E..
multiFLEX: Flexible Multi-utility, Multi-service Smart Metering Architecture for Energy Vectors with Active Prosumers.
DOI: 10.5220/0005483202880293
In Proceedings of the 4th International Conference on Smart Cities and Green ICT Systems (SMARTGREENS-2015), pages 288-293
ISBN: 978-989-758-105-2
Copyright
c
2015 SCITEPRESS (Science and Technology Publications, Lda.)
tributed architectures for Smart Grid. Section 3 de-
scribes the characteristics and objectives addressed by
multiFLEX. Section 4 introduces the proposed soft-
ware architecture and its expected impacts are pre-
sented in Section 5. Finally, Section 6 provides the
concluding remarks.
2 STATE OF THE ART ON
DISTRIBUTED
ARCHITECTURE FOR SMART
GRIDS
In a Smart Grid scenario, different middleware so-
lutions and service oriented architectures have been
proposed to enable a pervasive monitoring and man-
agement of the the grid itself for providing ser-
vices. Moreover in (Karnouskos, 2009; Warmer et al.,
2009), the authors focus on the importance of ex-
ploiting a Web Services approach to establish the in-
teroperability between Smart Home and Smart Grid
contexts. Also at building level, service oriented ar-
chitectures (Candido et al., 2009a; Candido et al.,
2009b) and middelware solutions (Stavropoulos et al.,
2013) have been proposed. Moreover, such solutions
provide a set of API (Application Programming In-
terfaces) in order to promote the integration and the
communication between Smart Buildings and Smart
Grids. However, this is necessary but still not enough.
Indeed, to provide services suitable for the whole
Smart Grid, such as demand response, a complete and
widespread overview of the grid itself is needed. Fol-
lowing this view in (Patti et al., 2014a), the authors
propose an architecture for integrating different data-
sources for increasing the energy efficiency in heating
distribution networks at district level.
In the power system scenario, Kim et al. present
a data-centric infrastructure that exploits the pub-
lish/subscribe communication paradigm to allow de-
centralized monitoring and management of the grid
itself (Kim et al., 2010). The GridStat middle-
ware (Tomsovic et al., 2005; Hauser et al., 2005;
Gjermundrod et al., 2009) is another solution for
enabling the communication across the devices in
power system. However, GridStat works with its own
closed and dedicated network infrastructure (Ger-
manus et al., 2010), which is incompatible with In-
ternet, so new routers must be deployed. Villa et
al. present the CoSGrid middleware for measuring
and controlling the electrical power of heterogeneous
Smart Grid infrastructures (Villa et al., 2011). It
exploits a remote method invocation and an event
notification approach to enable the communication.
In (Patti et al., 2014b), the authors present a dis-
tributed software infrastructure for general purpose
services in power systems. They reached the interop-
erability across heterogeneous devices and built their
software architecture exploiting the LinkSmart mid-
dleware
1
, which creates a secure peer-to-per network.
With respect to the presented solutions, we pro-
pose multiFLEX, a multi-service and multi-utility ar-
chitecture that aims at facilitating the access of mul-
tiple actors to relevant data to foster the spreading
of various innovative services. As the previous so-
lutions, it enables the communication between het-
erogeneous devices in the grid. In addition, multi-
FLEX offers a cloud-based infrastructure to collects,
analyse and provide energy information from differ-
ent meters. It is worth noting that multiFLEX features
are not strictly related to power systems but they are
opened also to other utilities such as water, heating
and gas.
3 PROPOSED APPROACH
The existing applications acknowledge that metering
infrastructure is an enabling technology that needs to
be coupled with innovative services to reach energy
management by means of rewards, automation and
information. In order to reinforce customers’ engage-
ment in achieving energy efficiency, number of utili-
ties already operate demand response and direct load
control to limit and shift the peak loads. However,
new services can be integrated focusing on more com-
plex technical applications, transparent for the end-
user but nevertheless with higher social impact. mul-
tiFLEX pursues the ambition of innovating this sce-
nario exploiting:
multi-service approach that uses information
coming from electric, water, gas, district heating
meters to provide general purpose services;
substation meters to improve fault tolerance and
demand response capabilities of the network, tak-
ing into account local electric storage and genera-
tion;
advanced Non-Intrusive Appliances Load Moni-
toring (NIALM) techniques to profile user be-
haviours and introducing a user signature of en-
ergy consumption/production regarding electric-
ity, gas, water and heating;
demand response algorithms that exploit informa-
tion about energy flows from the meters and the
NIALM profiles.
1
https://linksmart.eu/redmine
multiFLEX:FlexibleMulti-utility,Multi-serviceSmartMeteringArchitectureforEnergyVectorswithActiveProsumers
289
Figure 1: multiFLEX: flexible, multi-utility, multi-service metering architecture.
To achieve this, a software infrastructure to gather
data from heterogeneous data-sources and to perform
real-time data processing is needed. This can be
achieved with a cloud system capable of aggregating
and correlating such information. Finally, the result-
ing software infrastructure have to provide end-users
with feedbacks suggestions for optimizing energy us-
age.
In order to allow innovative services and utilities
for the different actors playing in a Smart Grid context
(e.g. ESCOs, DSOs, prosumers and customers) a flex-
ible multi-utility and multi-service metering architec-
ture is needed. As shown in Figure 1, multiFLEX
integrates off-the-shelf meters placed at the users
for electric, water, heating and gas metering. Such
heterogeneous meters directly communicate with a
building concentrator, which is in charge to enable
a bidirectional communication with the central cloud
system. For what concerns the electricity grid, mul-
tiFLEX also integrates off-the-shelf electric meters
deployed in MV/LV (Medium Voltage/Low Voltage)
substations. The central cloud system is in charge
of: i) collecting data from the building concentrators,
thus from the different meters at user home, and from
MV/LV substation meters; ii) post-processing incom-
ing information exploiting algorithms for data collec-
tion, fusion and mining; iii) providing a set of API
and tools for general purpose services and utilities.
Example of services for the prosumers are the access
to historical records of the energy consumption and
their analysis with saving suggestions. While, exam-
ple of services for the DSOs are: i) fault detection; ii)
detection of thefts; iii) demand response; iv) energy
storage integration.
In order to achieve a flexible multi-service and
multi-utility infrastructure, multiFLEX has to ensure
real-time bidirectional communication with each me-
ter in the Smart Grid. As shown in Figure 1, multi-
FLEX integrates heterogeneous meters (e.g. electric,
gas, water and heating meters) deployed at users and
also electric meters at MV/LV substations to monitor
the whole Smart Grid and to manage the loads in real-
time. Following this approach, the meters are seen as
a network device connected to internet. Hence, mul-
tiFLEX enables:
real-time readings management;
real-time accounting activities management;
real-time information to customers through a suit-
able interface structure;
detection of energy thefts;
near-real-time grid level and user level fault de-
tection allowing optimal alarming and first inter-
vention systems to be adopted;
demand response together with optimal integra-
tion of distributed generation and storage systems.
In order to enable a non intrusive automated en-
ergy monitoring systems to profile the energy con-
sumptions of the appliances, multiFLEX integrates in
its cloud a load profiling system that leverages the
NIALM technology (Zoha et al., 2012). NIALM is
a signal processing technique, which discerns the en-
ergy consumption of the appliances from the aggre-
gated data acquired from a single point of measure-
ment. It exploits transient electricity consumption
SMARTGREENS2015-4thInternationalConferenceonSmartCitiesandGreenICTSystems
290
Figure 2: multiFLEX software architecture layers.
events that are used to uniquely characterize each ap-
pliance. The resulting load profiles are used for i)
suggesting changing in user behaviours that will al-
low savings; ii) input for the demand response man-
agement system; iii) energy consumptions analysis.
4 multiFLEX SOFTWARE
ARCHITECTURE
In order to address the objectives described in Sec-
tion 3, we propose an innovative multi-service dis-
tributed infrastructure to allow remote management
and provide new services to DSOs, prosumers and fi-
nal customers. To achieve this, multiFLEX has to en-
able the interoperability across heterogeneous devices
exploiting middleware technologies that seem to be
promising along this direction (Patti et al., 2013; Patti
et al., 2014b). As shown in Figure 2, the resulting
architecture is organized in the following layers:
Integration Layer: it is in charge to enable the inter-
operability across the heterogeneous devices by
abstracting a certain technology to a Web Ser-
vices. Hence, it translates whatever kind of lan-
guage the low-level technology speaks into Web
Services.
M2M Layer: the Machine-to-Machine (M2M)
Layer is responsible for data communication
based on publish/subscribe approach (Eugster
et al., 2003). This approach increases the scalabil-
ity because it removes the interdependencies be-
tween producer and consumer of the information
allowing the development of services completely
independent from the systems and deployed de-
vices. Furthermore, it allows the development of
distributed applications and services that react in
real-time to certain events.
Storage Layer: it collects the data coming from the
meters and devices deployed across the city. mul-
tiFLEX is a modular and flexible infrastructure
where heterogeneous technology can be plugged
in. Hence, databases in the storage layer exploit a
non-relational schema-less approach.
Application Layer: it provides a set of API, tools
and distributed applications to manage and post-
process the information coming from the lower
layers.
Security Layer: it provides features to enable a
secure and trusted communication. It controls
whether a device or service can be trusted or not.
Therefore, it enables mutual authentication by
providing the means to create a public key infras-
tructure. Furthermore, it allows cryptographic op-
erations for message protection in order to guar-
antee the confidentiality between the parties.
In addition to the presented multiFLEX software
architecture layers, we identified two main platforms
for providing feedback and post-processed analysis to
end-users:
User Interface Platform: it is built on top of the Ap-
plication Layer. It is a user-friendly multi-service
platform that provides end users’ energy con-
sumption, tips and suggestions to promote green
behaviours, to reduce energy waste and related
costs as well. This platform is also used by stake-
holders, such as the multi-utility companies, in or-
der to manage and control deployed meters, for
instance to check system status, operating condi-
tions and faults.
NIALM Platform: It is a cloud-based data process-
ing unit dedicated to profile the user electrical en-
ergy consumptions. It exploits energy disaggre-
gation algorithm to discern the consumption of
the appliances from the aggregated data acquired
by the single meter at home. From this informa-
tion, the algorithm extracts the signature patterns
in order to associate each transient to a specific
appliance. Then the resulting disaggregated in-
formation are made available to provide energy
awareness (via the User Interface Platform) and
to forecast more accurately the energy demand in
the short term.
5 EXPECTED IMPACTS
In this section we analyse the expected impacts for the
multiFLEX architecture. From a preliminary analy-
sis, it might seem that the utility companies already
multiFLEX:FlexibleMulti-utility,Multi-serviceSmartMeteringArchitectureforEnergyVectorswithActiveProsumers
291
offers some of the proposed features such as: i) on-
line available user profiles; ii) on-line consumption
information; iii) suggestions for consumption opti-
mization; iv) basic predictions for future consump-
tion. However, the back-end block for data acquisi-
tion and processing is missing. Indeed, multiFLEX
aims on enabling a multi-service and multi-utility
platform in which different services like electricity,
water, gas and district heating, converge. Following
this view, multiFLEX provides end-users a complete
overview of their energy consumptions in real-time.
In addition, concerning the electrical consumption,
multiFLEX can analyse more accurately the users
load profile thanks to the NIALM platform avoiding a
massive deployment of sensor devices, such as smart
plugs. Therefore the impact will be important both on
the provider and consumer side, for instance: i) re-
ducing costs for the utility companies on reading the
data; ii) providing real-time services to end-users; iii)
detecting faults; iv) providing demand response.
The availability of real time and open data at ev-
ery level of energy distribution (e.g electricity, water,
gas and heating) chain will be the turning point for
promoting the active involvement of end-users and
fostering other working actors, such as energy man-
agers, ESCOs and aggregators, in providing innova-
tive services. No one of the players listed before
can really make the difference without analysing and
exploiting data with such a granularity as to show
consumers habits, electric devices status and faults.
Hence, enabling the interoperability and interconnec-
tion between different meters and multiFLEX archi-
tecture, that canbe considered also as a common data-
exchange platform, will foster the spreading of inno-
vative services.
Social impacts will be strictly related to real-time
data availability even at consumer level. Indeed the
knowledge of each own consumptions is the starting
point for other more integrated and innovative ser-
vices (e.g. remote house control systems) that will
change people behaviours.
Sharing information related to the development
and the status of the Smart Grid has to be the ob-
jectives for its continuous expansion, stimulating a
virtuous circle. DSO will benefit from a more effec-
tive operation and maintenance management systems
while consumers will benefit from information of en-
ergy bad habits and more tailored energy offers.
5.1 Impacts on End Users
Knowing the itemized consumption of the household
devices means having the necessary actionable feed-
back information to propose for reducing the energy
waste. multiFLEX provides dual benefits both to the
end-users and to the energy providers as well. In the
former case multiFLEX can provide detailed knowl-
edge of the household consumption for each appli-
ance and suggest personalized tips to reduce energy
waste. Generally, the benefits for end-users are sum-
marized in the following:
knowing the disaggregated energy of the house-
hold appliances;
discovering which appliance is the most ineffi-
cient by comparing its consumption with more ef-
ficient models present in other apartments;
being aware of consumption for each appliance in
terms of energy, money and CO
2
footprints. This
can help the end-user in taking positive decisions
to reduce the energy waste by 15-20% (Darby,
2006; Darby, 2008);
comparing the disaggregated appliance consump-
tion among different weeks, months or years;
observing the energy consumption in real-time by
mean of a smartphone applications to monitor the
apartment and receive alarms whenever the en-
ergy situation in the apartment is not as expected.
5.2 Impacts on Utility Providers and
Energy Operators
In view of the utility provider, multiFLEX can in-
crease the efficiency of demand response strategies by
keeping track of energy use patterns and behaviour of
their customers. Hence from the energy operator and
utility provider viewpoints, the main advantages are:
profiling consumer energy behaviours in order to
predict the short term energy demand;
offering personalized pricing policies to con-
sumers after profiling;
providing more efficient demand response strate-
gies to optimize the energy management during
peak periods balancing the consumers’ energy
loads (Bergman et al., 2011).
6 CONCLUSIONS
In this paper, we presented the characteristics and the
objectives of multiFLEX, a flexible multi-utility and
multi-service metering architecture for energy vectors
with active prosumers. Furthermore, we discussed
the benefits and the expected impacts that such ar-
chitecture can have in a Smart Grid context. In-
deed, multiFLEX integrates different meters into a
SMARTGREENS2015-4thInternationalConferenceonSmartCitiesandGreenICTSystems
292
distributed infrastructure with the aim of gathering,
post-processing and analysing heterogeneous infor-
mation from different meters and data-sources. Thus,
multiFLEX provides an overview of both energy con-
sumption and production in the grid from different
viewpoints, fostering working actors in promoting
new and innovative services.
ACKNOWLEDGEMENTS
This work was supported by Flexmeter, which is a
H2020 European Research Project.
REFERENCES
Bergman, D., Jin, D., Juen, J., Tanaka, N., Gunter, C.,
and Wright, A. (2011). Distributed non-intrusive load
monitoring. In IEEE PES ISGT.
Candido, G., Colombo, A., Barata, J., and Jammes, F.
(2009a). Service-oriented infrastructure to support the
deployment of evolvable production systems. IEEE
Trans. on Industrial Informatics.
Candido, G., Jammes, F., Barata, J., and Colombo, A.
(2009b). Generic management services for dpws-
enabled devices. In Proc of IEEE IECON.
Darby, S. (2006). The effectiveness of feedback on en-
ergy consumption: A review for defra of the litera-
ture on metering, billing and direct displays. Tech.
report,Environmental Change Institute, University of
Oxford.
Darby, S. (2008). Energy feedback in buildings: improv-
ing the infrastructure for demand reduction. Building
Research & Information.
Eugster, P. T., Felber, P. A., Guerraoui, R., and Kermarrec,
A.-M. (2003). The many faces of publish/subscribe.
ACM CSUR.
European Parliament (1999). Directive 96/92/ec concerning
common rules for the internal market in electricity. In
Second report to the Concil and the Europian Parl-
ment on harmonization requirements.
Germanus, D., Dionysiou, I., Gjermundrod, H., Khelil, A.,
Suri, N., Bakken, D., and Hauser, C. (2010). Leverag-
ing the next-generation power grid: Data sharing and
associated partnerships. In Proc. of IEEE PES.
Gjermundrod, H., Gjermundrod, H., Bakken, D., Hauser,
C., and Bose, A. (2009). Gridstat: A exible qos-
managed data dissemination framework for the power
grid. IEEE Trans. on Power Delivery.
Hauser, C., Bakken, D., and Bose, A. (2005). A fail-
ure to communicate: next generation communication
requirements, technologies, and architecture for the
electric power grid. IEEE Power and Energy Mag-
azine.
Karnouskos, S. (2009). The cooperative internet of things
enabled smart grid. In Proc. IEEE ISCE.
Kim, Y.-J., Thottan, M., Kolesnikov, V., and Lee, W. (2010).
A secure decentralized data-centric information in-
frastructure for smart grid. IEEE Communications
Magazine.
Patti, E., Acquaviva, A., and Macii, E. (2013). Enable sen-
sor networks interoperability in smart public spaces
through a service oriented approach. In 5th IEEE
IWASI.
Patti, E., Acquaviva, A., Sciacovelli, A., Verda, V., Martel-
lacci, D., Castagnetti, F., and Macii, E. (2014a). To-
wards a software infrastructure for district energy
management. In 12th IEEE EUC, pages 215–220.
Patti, E., Syrri, A., Jahn, M., Mancarella, P., Acquaviva,
A., and Macii, E. (2014b). Distributed software in-
frastructure for general purpose services in smart grid.
IEEE Trans. on Smart Grid, pages 1–8.
Stavropoulos, T. G., Gottis, K., Vrakas, D., and Vlahavas, I.
(2013). awesome: A web service middleware for am-
bient intelligence. Expert Systems with Applications.
Tomsovic, K., Bakken, D., Venkatasubramanian, V., and
Bose, A. (2005). Designing the next generation of
real-time control, communication, and computations
for large power systems. Proc. of the IEEE.
Villa, D., Martin, C., Villanueva, F., Moya, F., and Lopez, J.
(2011). A dynamically reconfigurable architecture for
smart grids. IEEE Trans. on Consumer Electronics.
Warmer, C., Kok, K., Karnouskos, S., Weidlich, A., Nes-
tle, D., Selzman, P., Ringelstein, J., Dimeas, A., and
Drenkard, S. (2009). Web services for integration of
smart houses in the smart grid. In Grid-Interop - The
road to an interoperable grid.
Zoha, A., Gluhak, A., Imran, M. A., and Rajasegarar, S.
(2012). Non-intrusive load monitoring approaches for
disaggregated energy sensing: A survey. In Sensors.
multiFLEX:FlexibleMulti-utility,Multi-serviceSmartMeteringArchitectureforEnergyVectorswithActiveProsumers
293