Distributed Infrastructure for Multi-Energy-Systems Modelling and

Co-simulation in Urban Districts

Lorenzo Bottaccioli, Edoardo Patti, Enrico Macii and Andrea Acquaviva

Department of Control and Computer Engineering, Politecnico di Torino, Italy

Keywords:

Multi-Energy-Systems, Simulation Infrastructure, Renewable Energy, Distributed Software Infrastructure,

Smart Grid, Distributed Systems, Distribution Network.

Abstract:

In recent years, many governments are promoting a widespread deployment of Renewable Energy Sources

(RES) together with an optimization of energy consumption. The main purpose consists on decarbonizing the

energy production and reducing the CO

footprints. However, RES imply uncertain energy production. To

foster this transition, we need novel tools to model and simulate Multi-Energy-Systems combining together

different technologies and analysing heterogeneous information, often in (near-) real-time. In this paper, ﬁrst

we present the main challenges identiﬁed after a literature review and the motivation that drove this research

in developing MESsi. Then, we propose MESsi, a novel distributed infrastructure for modelling and co-

simulating Multi-Energy-Systems. This infrastructure is a framework suitable for general purpose energy

simulations in cities. Finally, we introduce possible simulation scenarios that have different spatio-temporal

resolutions. Space resolution ranges from the single dwelling up to districts and cities. Whilst, time resolution

ranges from microseconds, to simulate the operational status of distribution networks, up to years, for planning

and refurbishment activities.

1 INTRODUCTION

Nowadays, one of the main challenges in our socie-

ties consists on reducing greenhouse gas emissions as

highlighted during the international conference on cli-

mate changes (United Nations, FCCC, 2015). Many

countries are investing on developing and deploying

Renewable Energy Source (RES) to reduce the depen-

dence on fossil fuels for energy generation. Moreover,

novel ICT (Information Communication Technology)

solutions can increase the demand ﬂexibility by ma-

naging the uncertain production of RES and by opti-

mizing the energy consumption in cities.

In this paper, we propose a distributed infra-

structure, called MESsi, for modelling and simula-

ting Multi-Energy-Systems (MES) by exploiting no-

vel ICT solutions, such as cyber-physical-systems,

Internet-of-Things (IoT), cloud computing and cog-

nitive computing. As pointed out by (Mancarella,

2014), an in-depth simulation and analysis of MES

is required to increase the ﬂexibility of energy sys-

tems by integrating different resources for both elec-

tric and thermal energy. Furthermore, ICT and MES

offer valid options to foster novel services for smart

This work was partially supported by the EU project FLEXMETER.

energy management. For example they can foster

events of Demand Response (DR) and Demand Side

Management (DSM) by integrating buildings equip-

ped with heat pumps, CHP (Combined Heat Power)

or HVAC (Heating, Ventilation and Air Conditioning)

systems (Molitor et al., 2014). The scope of this pa-

per consists on presenting the methodology and the

conceptual overview of MESsi, which is under deve-

lopment and only some modules have been validated.

The rest of the paper is organized as follows.

Section 2 presents motivations and challenges that

drove our research on developing such infrastructure.

Section 3 reviews relevant state of the art solutions

for modelling and simulating Multi-Energy-Systems.

Section 4 introduces the MESsi platform and possible

simulation scenarios. Finally, Section 5 discusses our

concluding remarks.

2 MOTIVATIONS AND

CHALLENGES

This research aims at developing a distributed

infrastructure to model and co-simulate Multi-

Energy-Systems (a.k.a. MESsi) in urban context.

262

Bottaccioli, L., Patti, E., Macii, E. and Acquaviva, A.

Distributed Infrastructure for Multi-Energy-Systems Modelling and Co-simulation in Urban Districts.

DOI: 10.5220/0006764502620269

In Proceedings of the 7th International Conference on Smart Cities and Green ICT Systems (SMARTGREENS 2018), pages 262-269

ISBN: 978-989-758-292-9

MESsi combines together different technologies and

heterogeneous information to model the energy ﬂows

and to simulate the impact of novel control strategies

in cities and distribution networks. It also can exploit

information coming in (near-) real-time from Inter-

net connected devices installed across the city. Furt-

hermore, MESsi provides features to simulate how

such novel policies affect the energy marketplace and

to analyse the effects and/or limitations of regula-

tory frameworks. On these premises, MESsi is an

infrastructure for simulations as a service that can

be used by different stakeholders to build and ana-

lyse new energy scenarios for short- and long-term

planning activities and for testing and managing the

operational status of Multi-Energy-Systems. Exam-

ples of scenarios that combine together thermal and

electricity trends (load and/or generation) to simulate

the energetic behaviour of buildings, districts and ci-

ties are: i) Installation of Renewable Energy Sources,

ii) Grid reconﬁguration, iii) Demand Response and

iv) Demand Side Management. To achieve this pur-

pose, MESsi needs to combine novel or already ex-

isting modelling and simulation tools together with

real-time simulator (e.g. OPAL-RT or RTDS). At

the same time, it needs to correlate heterogeneous

information, such as: i) measurements retrieved in

(near-) real-time from IoT devices deployed across

the city (e.g. information on multi-vector energy

trends, weather, indoor temperature in buildings, sta-

tus of the distribution grids); ii) Building Information

Models (BIM), grid models and Geographical Infor-

mation Systems (GIS); iii) topology of energy distri-

bution networks; iv) urban cartographies and informa-

tion on population censuses.

To realize the MESsi infrastructure, we identi-

ﬁed the following key challenges from a literature re-

view (Mancarella, 2014; Molitor et al., 2014; Alle-

grini et al., 2015; Keirstead et al., 2012; Van Beuze-

kom et al., 2015) that needs to be addressed:

i) Simulation of buildings dynamics: MESsi has to

provide features for analysing both thermal and elec-

trical dynamics in buildings. For example, model-

ling and simulating thermal dynamic includes also

the analysis of indoor temperature variations related

to power consumption. In this view, information on

thermal inertia and/or heat storages can be given as

input to control policies for shaving demand peaks in

district heating networks (Brundu et al., 2017; Verda

et al., 2016) or for DR and DSM if heating and cool-

ing systems are supplied by electric generators or

CHPs.

ii) Simulation of novel energy management policies:

Novel control policies needs to be evaluated in a re-

alistic environment before being applied in a real-

world context. Thus, the effects in terms of energy

efﬁciency, energy optimization, distribution network

reliability and economic value can be evaluated in-

depth.

iii) Simulation of distribution networks: MESsi must

be able to simulate the energy distribution network

to provide energy management policies with informa-

tion on the status of the network itself. For example,

from these simulations possible congestions, failures

and unbalances can be evaluated in a realistic scena-

rio. For this purpose, simulators like OPAL-RT and

RTDS need to be integrated in the infrastructure to

perform real-time simulations of the distribution net-

work with microseconds time-steps.

iv) Evaluation of RES impacts on the marketplace:

The impact of Renewable Energy Sources needs to be

evaluated to better design and apply novel control po-

licies and actions that affect the energy marketplace,

such as signal pricing and load balancing. Moreover,

these results can be analysed to better understand the

the limitation of current regulatory frameworks.

v) Simulations with different spatio-temporal resoluti-

ons: MESsi has to provide features to simulate energy

phenomena with different time and space resolutions.

Time resolution ranges from the microseconds, for

analysing the operational status of distribution sys-

tems, up to years, for planning and refurbishment acti-

vities. Whilst, space resolution ranges from the single

dwelling up to districts and cities.

vi) (Near-) real-time integration of real-world

information: Real-world information sent in

(near-) real-time by heterogeneous Internet con-

nected devices are needed to develop more accurate

event-based models for analysing the operational

status of the grid, for developing ad testing more

efﬁcient control policies and for planning and

refurbishment activities (Bottaccioli et al., 2017a).

vii) Modularity and extendibility in integrating data,

models and simulators: Modularity and extendibility

are two main features for Multi-Energy-Systems. In

particular, MESsi needs to be designed to integrate in

a plug-and-play fashion heterogeneous data-sources,

models and simulators. This makes the overall in-

frastructure suitable for simulating different energy

scenarios, becoming a general purpose framework for

energy simulations in cities. Modularity and extendi-

bility are also two main requirements to allow future

extensions with low cost and small architectural im-

pacts.

viii) Scalability of the infrastructure: Horizontal and

vertical scalability of the infrastructure is another key

requirement of MESsi. Indeed, it needs to scale up

quickly and easily because simulating a city or a dis-

trict implies the interaction of thousand of concurrent

Distributed Infrastructure for Multi-Energy-Systems Modelling and Co-simulation in Urban Districts

263

entities. This becomes critical if real-time simulations

of power distribution network must be performed.

3 STATE OF THE ART

In the last years, the study of Multi-Energy-Systems is

becoming crucial to de-carbonize energy production

and also to foster a widespread deployment of RES.

To achieve it, we need tools for an in-depth analysis

and simulation of MES for both electrical and thermal

energy (Mancarella, 2014). In (Molitor et al., 2014;

Allegrini et al., 2015; Keirstead et al., 2012; Van Beu-

zekom et al., 2015), authors present a complete over-

view of literature tools and models for MES analy-

sis. In this section, we report relevant state of the art

solutions on modelling and simulation platforms for

Multi-Energy-Systems, identifying challenges and li-

mitations as well.

DER-CAM (Firestone, 2004) is a useful tool for

planning and operational analysis of power distribu-

tion networks. It aims at providing guidelines for fu-

ture investment. The input can be given with a resolu-

tion up to 5 minutes. HOMER (Lambert et al., 2006)

helps on studying different micro-grid conﬁgurations

based on hourly input data. EnergyPLAN (Lund,

2011) is another solution useful for both operatio-

nal and planning activities. It receive input data up

to hourly values. However, none of these solutions

provides features for detailed power ﬂow analysis or

thermal simulations in buildings. Moreover, they are

not ﬂexible in integrating new scenarios in the simu-

lation process and they do not exploit data coming in

(near-) real-time from real devices installed across the

city.

GRIDSpice (Anderson et al., 2014) is a dis-

tributed platform that co-simulate power ﬂows and

data communication in smart-grid scenarios. It in-

tegrates third-party software like MATPOWER and

GridLAB-D to simulate power generation, demand

and distribution. It exploit a cloud-based architec-

ture to parallelize the computation of large scale mo-

dels. Also in this case, GRIDSpice neglects ther-

mal simulations in buildings and does not exploit

(near-) real-time information from real devices.

DIMOSIM (Riederer et al., 2015) is a platform

to perform MES simulations in urban districts. It

enables thermal simulations in buildings but it lacks

of electrical ﬂows simulations in power grids. MO-

SAIK (Sch

utte et al., 2011) is a distributed platform

for co-simulation of electrical ﬂows in smart grid sce-

narios. It provides an integration of their Matlab mo-

dels with PowerFactory, a third-party software, to ex-

ploit Photovoltaic (PV) and Load generation proﬁles.

IDEAS (Baetens et al., 2015) is an open source plat-

form based on Modelica modelling language. It co-

simulate Demand Side Management strategies where

thermal request of buildings affects power distribu-

tion networks. MESCOS (Molitor et al., 2014) is a

co-simulation platform for district energy systems. It

simulates Demand Response and Demand Side Ma-

nagement policies by integrating both electrical and

thermal loads. The main limitations of these solu-

tions are summarized as follows: i) they do not in-

tegrate (near-) real-time information from real devi-

ces; ii) they do not exploit a real-time simulator (e.g.

OPAL-RT and RTDS); iii) the integration with other

simulation tools is not easy. In addition to that, MO-

SAIK lacks in simulating thermal behaviours in buil-

dings.

In (Abrishambaf et al., 2017), authors present a

distributed platform for real-time co-simulation of

Demand Response events in microgrids. The platform

integrates the OPAL-RT simulator and exploits infor-

mation coming in real-time from real devices. Howe-

ver, the platform does not simulate thermal behavi-

ours in buildings.

HUES (Bollinger and Evins, 2015) platform aims

at facilitating the integration of different models for

MES analysis. It implements a repository layer that

includes all the platform modules whose functiona-

lities are described in a semantic wiki. However,

HUES neglects on an interconnection among the plat-

form’s modules and lacks on integrating data coming

in (near-) real-time from devices installed across the

city.

In our previous work (Bottaccioli et al., 2017b),

we presented a real-time architecture for co-

simulation of novel control policies in smart grid with

RES. In its core, it leverages upon an OPAL-RT si-

mulator and exploits real-time information from real

devices. However, this solution neglects in simulating

thermal behaviours in buildings.

With respect to literature solutions, we propose

MESsi, a distributed infrastructure for modelling and

simulating Multi-Energy-Systems. It aims at overco-

ming the highlighted limitations and addressing the

main challenges identiﬁed in Section 2 to evaluate

general purpose simulation scenarios. In particular,

MESsi performs simulations for both thermal and

electrical distribution networks with different spatio-

temporal resolutions. It exploits the OPAL-RT real-

time simulator that allows in-depth simulations with

microseconds time-steps. It provides features to per-

form detailed power ﬂow analysis, thermal simulati-

ons in buildings and evaluation of RES impacts on the

marketplace. Furthermore, MESsi integrates data co-

ming in (near-) real-time from real devices installed

SMARTGREENS 2018 - 7th International Conference on Smart Cities and Green ICT Systems

264

across the city. Finally, it eases the interconnection

among MESsi components and third-party models

and simulators in a plug-and-play fashion.

4 INFRASTRUCTURE FOR MES

MODELLING AND

CO-SIMULATION

In this section, we present MESsi, a distributed infra-

structure for real-time modelling and co-simulation of

Multi-Energy-Systems in cities (see Figure 1). This

infrastructure exploits the microservice design pat-

tern (Fowler and Lewis, 2014; Newman, 2015) to in-

crease both scalability and extendibility of the system,

and to ease its maintenance. Indeed, the microservice

approach deﬁnes software architecture as a set of loo-

sely coupled and collaborating services. Thus, our so-

lution is ﬂexible in modelling and co-simulating dif-

ferent energy ﬂows in a single solution made of diffe-

rent interoperable components or modules that can be

deployed in a plug-and-play fashion.

The proposed infrastructure consists of ﬁve lay-

ers. From left to right in Figure 1, both Environ-

mental and Physical Layers includes the heteroge-

neous data-sources needed by the different compo-

nents in the system. The Cyber Layer enables the

communication among the different modules in the

ﬁve layers by exploiting either the request/response or

publish/subscribe (Eugster et al., 2003) communica-

tion paradigms. The Modelling and Simulation Layer

consists of different components that simulate energy

phenomena and multi-energy-ﬂows. Finally, the Si-

mulation Scenarios Layer provide end-users with a

set of tools and API (Application Programming Inter-

faces) to build and run their MES simulation scena-

rios. At this layers, end-users can easily access to all

the information made available by the modules in the

previous layers.

4.1 Data Sources

The proposed solution integrates heterogeneous data-

sources needed by the simulation components. In par-

ticular we group them in two layers, Environmental

Layer and Physical Layer (see Figure 1).

The Environmental layer integrates all the infor-

mation needed to describe a city. Among the others

this layer includes:

i) Geographical Information Systems (GIS) integrate

georeferenced information about the different entities

(e.g. devices, buildings and pipelines) in cities. It

also includes cartographies cadastral maps and Digi-

tal Elevation Models.

ii) Building Information Models (BIM) are parametric

3-Dimensional models, where each model describes a

building, both structurally and semantically.

iii) System Information Models (SIM) describe size

and structure of energy distribution networks. SIM is

built by exploiting parametric and topological data.

iv) Weather Data are retrieved by third party servi-

ces, such as (Weather Underground, 2017). This in-

formation is georeferenced and collected by personal

weather stations deployed in cities.

v) Census data are information about different cha-

racteristics and behaviours of citizens, such as popu-

lation distribution, dwelling size and appliances dis-

tribution.

On the other hand, the Physical layer integrates

data coming from physical systems and Internet con-

nected devices in (near-) real-time. Among the others

this layer includes:

i) Measurements of energy production from Distribu-

ted Generation.

ii) Status of Distribution Grid that are needed to simu-

late energy ﬂows and evaluate the integration of RES.

Thus, information sampled by devices monitoring the

energy distribution network.

iii) Information sent by IoT devices, such as Ambient

sensors, multi-vector Smart Meters (i.e. electricity,

gas, heating and water) and Actuators.

4.2 Data Communication

The Cyber layer is in charge of enabling data ex-

change among the different components in our in-

frastructure. It exploits both the synchronous and

asynchronous communication paradigms adopting

both request/response and publish/subscribe (Eug-

ster et al., 2003) approaches, respectively. Re-

quest/response allows a fast bidirectional commu-

nication to send/access information to/from diffe-

rent components in our infrastructure (either har-

dware or software), using, for instance, REST Web

Services (Fielding and Taylor, 2002). Whilst, pu-

blish/subscribe is complementary to request/response

and allows (near-) real-time data transmission. Pu-

blish/subscribe removes the interdependencies bet-

ween producer and consumer of information. This al-

lows developers in creating distributed software com-

ponents that are independent from data-sources and

can react in (near-) real-time to certain events. Thus,

publish/subscribe enables a data-driven and event-

based communication that also increases the scalabi-

lity of the system as pointed out in (Patti et al., 2016).

In the proposed solution, we adopted MQTT proto-

Distributed Infrastructure for Multi-Energy-Systems Modelling and Co-simulation in Urban Districts

265

Figure 1: Schema of the proposed MESsi infrastructure.

col (MQTT, 2017), which is an implementation of

publish/subscribe.

As shown in Figure 1, the Cyber layer consists

of three main modules the Communication Adapter,

the Data Integration Platform and Smart Metering In-

frastructure. The Communication Adapter enables

the interoperability across the heterogeneous devi-

ces in the Physical Layer. Whilst, the Data Inte-

gration Platform integrates third party software and

platforms in the Environmental Layer. Both act as a

bridge between the components of infrastructure and

the underlying technologies, either hardware or soft-

ware. In this view, each technology needs a spe-

ciﬁc Communication Adapter or a Data Integration

Platform to provide common and uniﬁed interfaces to

access low-level functionalities through REST Web

Services and/or MQTT. Thus, both Communication

Adapter and Data Integration Platform are key com-

ponent to access each low-level technology transpa-

rently. Finally, MESsi provides features to integrate

also third party Smart Metering Infrastructure, such

as (Pau et al., 2017), that makes available histori-

cal data collected from real distribution networks and

post-processed information output of its services.

4.3 Modelling and Simulation

The Modelling and Simulation Layer, in Figure 1,

consists of different software components to simu-

late environmental conditions (green boxes), electri-

cal energy (light blue boxes) and thermal energy (yel-

low boxes).

The Solar Radiation Decomposition is a software

module that decompose Global Horizontal radiation

(GHI) into Direct Normal Incident radiation (DNI)

and Diffuse Horizontal Incident radiation (DHI) by

applying mathematical models such as (Ruiz-Arias

et al., 2010). The inputs are meteorological informa-

tion retrieved by Weather Data module in the Envi-

ronmental Layer. Often, weather stations sample only

GHI. Thus, this module is crucial because DNI and

DHI are needed to evaluate the solar heating gains and

to simulate incident solar radiation on tilted surfaces

(e.g. buildings’ rooftops).

The Rooftop Solar Radiation (Bottaccioli et al.,

2017c) module exploits GIS cartographies, real Weat-

her Data and results from Solar Radiation Decompo-

sition module to simulate incident solar radiation on

rooftops. Simulations are done in real-sky conditions

with a resolution of 15 minutes. It is able to detect

roof encumbrance (e.g. chimneys and dormers) and

to estimate their shadowing effects.

The Photovoltaic Energy module exploits the met-

hodology described in (Bottaccioli et al., 2017c). It

exploits both Rooftop Solar Radiation and Weather

Data modules to estimate the incident solar radiation

and the effects of the air temperature on the efﬁciency

of PV arrays. By exploiting GIS cartographies, it

also identiﬁes the suitable areas for PV deployment

on rooftops and simulates the energy production with

a resolution of 15 minutes.

The Smart Energy Management module offers an

environment fully integrated in MESsi to simulates

novel control policies for energy optimization suita-

ble for battery management, Demand Response and

Demand Side Management.

The Agent-Based Model for Market Impacts mo-

dule simulates the impact of RES (e.g. PV arrays) and

novel control policies on the electrical marketplace.

It aims at evaluating the role of emerging energy ag-

gregators to better understand the feasibility of such

actions in residential sectors from both regulatory and

economical viewpoints.

The Smart Grid Simulator module integrates a

Real-Time Simulators (e.g. Opal-RT or RTDS) as

depicted in (Bottaccioli et al., 2017b). It simulates

power distribution networks with different time re-

solutions ranging from microseconds to hours. Ex-

SMARTGREENS 2018 - 7th International Conference on Smart Cities and Green ICT Systems

266

ploiting the Communication adapters in the Cyber

layer, it is able: i) to access information from IoT

devices deployed across the real distribution network

in (near-) real-time and ii) to exchange data with the

Smart Energy Management, the Photovoltaic Energy

and Household Electricity Behaviour modules. The

Smart Grid Simulator module enables a more accu-

rate analysis of distribution network when different

control strategies are applied. Moreover, this module

can be used together with the Agent-Based Model for

Market Impacts module to simulate congestions and

unbalances, and to evaluate the electrical implication

of Demand Response events on distribution networks.

The Household Electricity Behaviour module re-

produces realistic electrical consumptions in residen-

tial houses. It builds a virtual model of households

occupancy and their activity patterns by exploiting

a Markov chain Monte Carlo simulation. The input

needed by this module are the information given by

Census Data (i.e. statistical information on the distri-

bution of population and appliances). In particular, it

uses results of the last national census that provides

statistical information on households occupancy and

citizens activities over the day. Once the virtual mo-

del is built, it translates these information into elec-

trical usages for each appliance in each single virtual

home. This module can build scenarios involving a

single house up to a full city. During the execution

of the simulation, the results are continuously made

available to the other modules of MESsi through the

Cyber Layer.

The District Heating Simulator provides an envi-

ronment fully integrated in MESsi to simulates no-

vel control policies for Heating Distribution Networks

(HDN) accounting also for their impacts on building

comfort (Brundu et al., 2017; Verda et al., 2016). Ad-

ditionally, this module provides tools to analyse and

predict the thermal behaviour of buildings connected

to HDN exploiting the KPIs and the methodology des-

cribed in (Acquaviva et al., 2015).

The Thermal Building Simulator follows the met-

hodology described in (Bottaccioli et al., 2017a). It

provides tools to simulate and analyse the thermal be-

haviour of buildings. It combines information about

BIM and GIS together with real Weather data and

environmental information coming from IoT Devices

deployed in the corresponding real buildings. This

module allows: i) (near-) real-time visualisation of

energy consumptions in buildings; ii) simulation of

indoor temperature trends and iii) evaluations of buil-

ding performances through energy models.

The Solar Thermal Collector module simulates

the behaviours of solar thermal panels in heating wa-

ter. It needs as input Weather Data and the results

from the Rooftop Solar Radiation.

4.4 Simulation Scenarios

The Simulation Scenarios Layer, the last in our infra-

structure (see Figure 1), provides end-users with a set

of tools and API to build and run different energy sce-

narios. This layer allows end-users to easily access

to information made available by the other module

in MESsi. Among the others this layer includes si-

mulation scenarios described in this section for which

MESsi as been designed.

Simulations can be performed to evaluate the ope-

rational states of distribution networks. For example,

events of DR and DSM in a urban MES can be analy-

sed taking into account RES (e.g. PV arrays) and buil-

dings equipped with electrical HVAC. These scena-

rios involve different MESsi modules: i) Smart Grid

Simulator to simulate the power network: ii) Photo-

voltaic Energy to estimate generation proﬁle of PV

systems; iii) Thermal Building Simulator to ana-

lyse the indoor temperature trends and to avoid dis-

comfort in case HVAC is switched off. Furthermore,

the Smart Metering Infrastructure can be used to re-

trieve historical data and to schedule DR and DSM

events that also involve common appliances (i.e. wa-

shing machine, dishwasher and boilers). Otherwise, if

real data are not available, Household Electricity Be-

haviour generates realistic load proﬁles for different

dwellings. These simulations span different spatio-

temporal resolutions at the same time. Simulations of

indoor temperature trends involve the single dwelling

or building with 15 minutes time resolution. Whilst,

grid behaviour simulations involve the whole district

or city with microsecond or second time resolution.

Thermal peak shaving is another scenario to eva-

luate the operational states of HDN. In this case, an

high request of hot water from buildings connected

to HDN causes a power peak at the thermal power

plant. This issue is normally managed with heat stora-

ges or with additional gas heaters that are used only

during this peak period. To avoid it, control policies

can be tested to shift in time the request of each single

building (Verda et al., 2016). This scenario involves

the Thermal Building Simulator to evaluate the indoor

temperature trends in each building and the District

Heating Simulator to evaluate the peak shaving. Both

modules needs 15 minutes time resolution.

As mentioned in the previous sections, MESsi can

be used also for strategic planning activities. For

example, it can simulate scenarios that involves the

Energy Aggregators, a new rising actors in the elec-

trical marketplace. In this case, Energy Aggregators

exploit the proposed infrastructure to analyse the cus-

Distributed Infrastructure for Multi-Energy-Systems Modelling and Co-simulation in Urban Districts

267

tomers participation and the effects of regulatory fra-

meworks in DR/DSM events. Consequently, Energy

Aggregators can evaluate their impact on the market-

place. This scenario is a long-term planning activity.

Thus, simulations needs monthly or yearly time reso-

lution. This scenario exploits the following modules:

i) Agent-Based Model for Market Impacts, ii) Hou-

sehold Electricity Behaviour , iii) Smart Metering In-

frastructure, iv) Photovoltaic Energy, v) Smart Grid

Simulator and vi) Thermal Building Simulator.

City managers can use MESsi to evaluate the so-

lar potential and its impact on the distribution net-

work considering also load proﬁles and network con-

strains. This scenario involves the following modules:

i) Photovoltaic Energy to estimate the generation pro-

ﬁle for each PV system installed in building rooftops;

ii) Smart Metering Infrastructure to retrieve real load

proﬁles or Household Electricity Behaviour to gene-

rate realistic energy consumption patterns, iii) Solar

Thermal Collector to simulate the behaviours of so-

lar thermal panels in heating water for domestic use

or for heat pumps. If the simulation scenario inclu-

des building heating systems supplied by solar ther-

mal panels, the Thermal Building Simulator is needed

to evaluate the impact on indoors temperature beha-

viours. The scenarios needs monthly or yearly time

resolution.

MESsi can also be used for testing or validating

already existing algorithms, such as Non-Intrusive

Load Monitoring (NILM). NILM is a signal proces-

sing technique, which discerns the energy consump-

tion of the appliances from the aggregated data acqui-

red from a single point of measurement, i.e. the Smart

Meter (Zoha et al., 2012). In this case, the input nee-

ded by the NILM algorithm are historical households

load proﬁles retrieved from Smart Metering Infra-

structure. As an alternative, the NILM service can ex-

ploit the Household Electricity Behaviour module to

create realistic electrical consumption patterns. Time

resolution for this scenario ranges from microseconds

to 1 second.

It is worth noting that, thanks to the microservice

design pattern, MESsi is opened to build and run new

simulation scenarios to meet latest requirements from

the end-users.

5 CONCLUSION

In this paper, we presented MESsi, which is a no-

vel distributed infrastructure for modelling and co-

simulating Multi-Energy-Systems in cities. First, we

discussed the motivations and challenges we addres-

sed to design such infrastructure. Then, we intro-

duced our proposed framework that is suitable for

general purpose energy simulations with different

spatio-temporal resolutions. MESsi combines dif-

ferent technologies and correlates heterogeneous in-

formation, also sent in (near-) real-time, to simu-

late multi-energy-ﬂows and to evaluate the impact

of novel policies in cities and distribution networks.

Finally, we discussed possible simulation scenarios

i) for analysing the operational status of energy dis-

tribution systems, ii) for planning and refurbishment

activities, and iii) for testing or validating already ex-

isting algorithms.

REFERENCES

Abrishambaf, O., Faria, P., Gomes, L., Sp

ınola, J., Vale, Z.,

and Corchado, J. M. (2017). Implementation of a real-

time microgrid simulation platform based on centrali-

zed and distributed management. Energies, 10(6):806.

Acquaviva, A., Apiletti, D., Attanasio, A., Baralis, E., Bot-

taccioli, L., Castagnetti, F. B., Cerquitelli, T., Chi-

usano, S., Macii, E., Martellacci, D., et al. (2015).

Energy signature analysis: Knowledge at your ﬁnger-

tips. In Big Data (BigData Congress), 2015 IEEE In-

ternational Congress on, pages 543–550. IEEE.

Allegrini, J., Orehounig, K., Mavromatidis, G., Ruesch,

F., Dorer, V., and Evins, R. (2015). A review of

modelling approaches and tools for the simulation of

district-scale energy systems. Renewable and Sustai-

nable Energy Reviews, 52:1391–1404.

Anderson, K., Du, J., Narayan, A., and El Gamal, A. (2014).

Gridspice: A distributed simulation platform for the

smart grid. IEEE Transactions on Industrial Informa-

tics, 10(4):2354–2363.

Baetens, R., De Coninck, R., Jorissen, F., Picard, D., Hel-

sen, L., and Saelens, D. (2015). Openideas-an open

framework for integrated district energy simulations.

In Proceedings of Building Simulation 2015.

Bollinger, L. A. and Evins, R. (2015). Facilitating model

reuse and integration in an urban energy simulation

platform. Procedia Computer Science, 51:2127–2136.

Bottaccioli, L., Aliberti, A., Ugliotti, F., Patti, E., Osello,

A., Macii, E., and Acquaviva, A. (2017a). Building

energy modelling and monitoring by integration of iot

devices and building information models. In Com-

puter Software and Applications Conference (COMP-

SAC), 2017 IEEE 41st Annual, volume 1, pages 914–

922. IEEE.

Bottaccioli, L., Estebsari, A., Pons, E., Bompard, E., Macii,

E., Patti, E., and Acquaviva, A. (2017b). A ﬂexible

distributed infrastructure for real-time co-simulations

in smart grids. IEEE Transactions on Industrial Infor-

matics.

Bottaccioli, L., Patti, E., Macii, E., and Acquaviva, A.

(2017c). Gis-based software infrastructure to model

pv generation in ﬁne-grained spatio-temporal domain.

IEEE Systems Journal.

SMARTGREENS 2018 - 7th International Conference on Smart Cities and Green ICT Systems

268

Brundu, F. G., Patti, E., Osello, A., Del Giudice, M., Ra-

petti, N., Krylovskiy, A., Jahn, M., Verda, V., Guelpa,

E., Rietto, L., and Acquaviva, A. (2017). Iot software

infrastructure for energy management and simulation

in smart cities. IEEE Transactions on Industrial In-

formatics, 13(2):832–840.

Eugster, P. T., Felber, P. A., Guerraoui, R., and Kermarrec,

A.-M. (2003). The many faces of publish/subscribe.

ACM CSUR.

Fielding, R. T. and Taylor, R. N. (2002). Principled design

of the modern web architecture. ACM Transactions on

Internet Technology, 2(2):115–150.

Firestone, R. (2004). Distributed energy resources custo-

mer adoption model technology data. Berkeley Lab,

Berkeley, CA, USA Case Study.

Fowler, M. and Lewis, J. (2014). Microservices. Available:

http://martinfowler.com/articles/

microservices.html.

Keirstead, J., Jennings, M., and Sivakumar, A. (2012). A

review of urban energy system models: Approaches,

challenges and opportunities. Renewable and Sustai-

nable Energy Reviews, 16(6):3847–3866.

Lambert, T., Gilman, P., and Lilienthal, P. (2006). Micro-

power system modeling with homer. Integration of

alternative sources of energy, pages 379–418.

Lund, H. (2011). Energyplan-advanced energy systems ana-

lysis computer model. Documentation version, 9.

Mancarella, P. (2014). Mes (multi-energy systems): An

overview of concepts and evaluation models. Energy,

65:1–17.

Molitor, C., Gross, S., Zeitz, J., and Monti, A. (2014). Mes-

cosa multienergy system cosimulator for city district

energy systems. IEEE Transactions on Industrial In-

formatics, 10(4):2247–2256.

MQTT (Accessed on November 2017). Available:

http://mqtt.org.

Newman, S. (2015). Building Microservices. O’Reilly Me-

dia, Inc.

Patti, E., Syrri, A. L. A., Jahn, M., Mancarella, P., Acqua-

viva, A., and Macii, E. (2016). Distributed software

infrastructure for general purpose services in smart

grid. IEEE Transactions on Smart Grid, 7(2):1156–

1163.

Pau, M., Patti, E., Barbierato, L., Estebsari, A., Pons, E.,

Ponci, F., and Monti, A. (2017). A cloud-based smart

metering infrastructure for distribution grid services

and automation. Sustainable Energy, Grids and Net-

works.

Riederer, P., Partenay, V., Perez, N., Nocito, C., Trigance,

R., and Guiot, T. (2015). Development of a simu-

lation platform for the evaluation of district energy

system performances. In Fourteenth International

IBPSA Conference, pages 2499–2506.

Ruiz-Arias, J., Alsamamra, H., Tovar-Pescador, J., and

Pozo-V

azquez, D. (2010). Proposal of a regressive

model for the hourly diffuse solar radiation under all

sky conditions. Energy Conversion and Management,

51(5):881–893.

Sch

utte, S., Scherfke, S., and Tr

oschel, M. (2011). Mo-

saik: A framework for modular simulation of active

components in smart grids. In Smart Grid Modeling

and Simulation (SGMS), 2011 IEEE First Internatio-

nal Workshop on, pages 55–60. IEEE.

United Nations, FCCC (2015). Adoption of the Paris

Agreement. Proposal by the President. Available:

http://unfccc.int/resource/docs/2015/cop21/eng/

l09r01.pdf.

Van Beuzekom, I., Gibescu, M., and Slootweg, J. (2015).

A review of multi-energy system planning and opti-

mization tools for sustainable urban development. In

PowerTech, 2015 IEEE Eindhoven, pages 1–7. IEEE.

Verda, V., Guelpa, E., Sciacovelli, A., Acquaviva, A., and

Patti, E. (2016). Thermal peak load shaving through

users request variations. International Journal of

Thermodynamics, 19(3):168–176.

Weather Underground (Accessed on November 2017).

Available: http://www.wunderground.com/.

Zoha, A., Gluhak, A., Imran, M. A., and Rajasegarar, S.

(2012). Non-intrusive load monitoring approaches

for disaggregated energy sensing: A survey. Sensors,

12(12):16838–16866.

Distributed Infrastructure for Multi-Energy-Systems Modelling and Co-simulation in Urban Districts

269