Assessment of energy sharing coefficients under the new Portuguese renewable energy communities regulation

Abstract

The recent published European legal framework (Directives (EU) 2018/2001 and 2019/944) on renewable energy consumption, and its Portuguese transposition (the Decree-Law 15/2022), opens the possibility for buildings to operate as energy communities. One of the objectives is to increase the use of locally generated energy from renewable sources, by sharing available surplus among participants, using sharing coefficients defined by the entire community.
Taking the actual legal framework into consideration, this paper presents an analysis of the energy sharing coefficients proposed by the newly published Portuguese legislation via the assessment of a renewable energy community, formed by public buildings, whose operation varies according to different sharing coefficient applied. Results show that time-variable energy sharing coefficients are the best option to the considered renewable energy community. Collected results also show that larger consumers can extract higher benefits from being integrated on a renewable energy community. These benefits decrease when buildings are allowed to self-consume local generated energy prior to the sharing process as demand inequalities become less important for the computation of the considered sharing coefficients. The entire community also presents better performance in this case.

Full text

Heliyon 9 (2023) e20599
Available online 7 October 2023
2405-8440/© 2023 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Assessment of energy sharing coefcients under the new
Portuguese renewable energy communities regulation
Humberto Queiroz
a
,
b
,
*
, Rui Amaral Lopes
a
,
b
, Jo˜
ao Martins
a
,
b
, Filipe Neves Silva
c
,
Luís Fialho
d
, Nuno Bilo
e
a
NOVA School of Science and Technology, NOVA University Lisbon, Portugal
b
Center of Technology and Systems (UNINOVA-CTS) and Associated Lab of Intelligent Systems (LASI), Portugal
c
EDP New, Portugal
d
University of ´
Evora, Renewable Energies Chair, Portugal
e
´
Evora Municipality, Portugal
ARTICLE INFO
Keywords:
Renewable energy community
Self-consumption
Energy sharing
Energy management
ABSTRACT
The recent published European legal framework (Directives (EU) 2018/2001 and 2019/944) on
renewable energy consumption, and its Portuguese transposition (the Decree-Law 15/2022),
opens the possibility for buildings to operate as energy communities. One of the objectives is to
increase the use of locally generated energy from renewable sources, by sharing available surplus
among participants, using sharing coefcients dened by the entire community.
Taking the actual legal framework into consideration, this paper presents an analysis of the
energy sharing coefcients proposed by the newly published Portuguese legislation via the
assessment of a renewable energy community, formed by public buildings, whose operation
varies according to different sharing coefcient applied. Results show that time-variable energy
sharing coefcients are the best option to the considered renewable energy community. Collected
results also show that larger consumers can extract higher benets from being integrated on a
renewable energy community. These benets decrease when buildings are allowed to self-
consume local generated energy prior to the sharing process as demand inequalities become
less important for the computation of the considered sharing coefcients. The entire community
also presents better performance in this case.
1. Introduction
The European Union (EU) has been encouraging its member states to increase the share of renewable energy sources (RES) into
their energy pool. This encouragement aims, among other goals, to mitigate climate changes effects, mostly caused by the emission of
greenhouse gases (GHG) into the atmosphere when energy is generated using fossil fuels, and to increase energy security [1]. One of
the main characteristics of the different RES technologies is the low level, or even the absence, of GHG emissions during the operation
phase. However, most RES exhibit a variable generation pattern, following the availability of the respective primary energy resource,
which increases the complexity of power systems management [2].
* Corresponding author. NOVA School of Science and Technology, NOVA University Lisbon, Portugal.
E-mail addresses: h.queiroz@campus.fct.unl.pt (H. Queiroz), rm.lopes@fct.unl.pt (R.A. Lopes), jf.martins@fct.unl.pt (J. Martins), lipe.
nevessilva@edp.pt (F.N. Silva), laalho@uevora.pt (L. Fialho), nuno.choraobilo@cm-evora.pt (N. Bilo).
Contents lists available at ScienceDirect
Heliyon
journal homepage: www.cell.com/heliyon
https://doi.org/10.1016/j.heliyon.2023.e20599
Received 1 June 2023; Received in revised form 5 September 2023; Accepted 30 September 2023

Heliyon 9 (2023) e20599
2
On the consumer side, the energy generated from local renewable sources is used to reduce energy import from the grid and the
respective costs [3]. To enhance the local usage of RES and associated benets related to self-consumption, the concepts of Renewable
Energy Community (REC) and Citizen Energy Community (CEC) have been introduced and encouraged by the EU and national entities
in the last years, as described in Section 1.1. These entities are entitled to share locally generated energy among their members at a
lower cost, when compared to the price of energy imported from distribution grids. This sharing process represents the cornerstone of
the referred energy communities as it offers real benets to small consumers and contributes to the engagement of citizens in topics
related to the energy sector, including renewable energy generation and self-consumption. By increasing the number of renewable
energy communities and citizen energy communities, members states can increase the use of RES and ultimately decrease the GHG
emissions of the energy sector, as aimed by the EU [1].
1.1. Self-consumption regulation
Self-consumption of RES refers to the immediate usage of energy generated from renewable sources to satisfy energy demand
within the premises where the respective energy conversion systems, commonly based on photovoltaic (PV) technology, are installed.
This practice can be conducted at both individual and aggregated levels and the literature shows that the second option presents better
results when compared to an individual operation due to the sharing of generation surplus among residential community members and
the respective reduction of energy import from distribution grids [4].
The European Commission published two important directives to foster the implementation of these activities, namely, Directive
(EU) 2018/2001 [5], published on December 11th, and Directive (EU) 2019/944 [6], published on May 30th. The former set goals for
the increasing of renewable energy sources usage. The latter, concerns electricity markets operation, considering common rules of
internal electricity markets. Both legal instruments also incentivise the association of consumers, residential, commercial, or
Nomenclatures
Acronyms
Bld Building
CEC Citizen energy community
CSME Collective self-consumption entity manager
DSO Distribution system operator
EU European Union
EV Electric vehicle
FEC Fixed and equal sharing coefcient
FPC Fixed and proportional sharing coefcient
GHG Greenhouse gases
H2020 Horizon 2020
NOCT Nominal operating cell temperature
P2P Peer-to-peer
PV Solar photovoltaic
REC Renewable energy community
RES Renewable energy source
STC Standard test conditions
VSC Variable sharing coefcient
ZCC Zero-carbon community
Variables
PG PV system instant power output [W]
N Number of modules in a PV array
A Useful area of a PV module [m
2
]
G Instantaneous global solar irradiance on the horizontal plane [W/m
2
]
η
c Instantaneous PV module’s efciency
η
i Efciency considering balance-of-system losses
η
STC PV module’s efciency at Standard Test Conditions
μ
PV module’s temperature coefcient [◦C
-1
]
θa Ambient temperature [◦C]
Tc,STC PV cell temperature at Standard Test Conditions [◦C]
Tc,NOCT PV cell temperature at Nominal Operating Cell Temperature [◦C]
θa,NOCT Ambient temperature at Nominal Operating Cell Temperature [◦C]
GNOCT Global solar irradiance at Nominal Operating Cell Temperature [W/m
2
]
BD Building demand [W]
H. Queiroz et al.

Heliyon 9 (2023) e20599
3
industrial, to trade their exibility and self-generated energy into local markets. The REC and CEC concepts are developed to comply
with this incentive. These two types of communities, despite their similarities regarding member participation and entitlement for
energy trading, have some specicities that differentiate them. The main difference concerns the fact that a CEC can own, establish,
purchase or lease electricity distribution networks (these possibilities are not available for a REC). The term REC is chosen in the
remaining of this paper for the sake of readability given that the referred difference between CEC and REC is not addressed in this
study.
Portugal, as one of the EU member states, addresses the importance of the aforementioned legal framework by the publication of
the Decree-Law 15/2022 [7], which approves the applicable legal regime for renewable energy self-consumption (both individual and
collective). Furthermore, Regulation 373/2021 was approved by the national Energy Services Regulatory Authority, providing further
details to regulate electricity self-consumption, both individual and collective through the transposition to the REC concept to the
Portuguese legal framework, and the relationship between the various stakeholders involved in this process [8].
1.2. Literature review
The collective association of consumers and/or prosumers to conduct collective self-consumption, instead of individual self-
consumption, is assessed in a variety of ways in the literature. These associations are implemented and studied to achieve different
objectives, such as energy savings and reduction of GHG emissions [9], higher PV self-consumption ratios [10], or improvement of
wellbeing on a grid fault via energy exibility [11]. For low-voltage grids, the deployment of energy communities can also offer
real-time scheduling of exibility [12]. Regarding energy carriers, electricity is not the only one possible to be used by REC members.
Authors in Ref. [13] described the analysis of a renewable polygeneration system connected to a district heating and cooling network
in Naples, Italy. The study incorporates thermodynamic, economic, and environmental factors, considering geothermal energy as the
primary source.
A signicant part of an energy community management framework regards the sharing of locally generated energy among the
participants. In Ref. [14], a sharing model with price-based demand response was developed, by a supply and demand ratio of shared
PV energy, while the respective consumption exibility of each one of the considered prosumers is used to develop an equivalent cost
model. The authors in Ref. [15] developed an entire business model for an energy community, where one of the considered aspects is
the fair distribution of the reward achieved by the community members. In this case study, community members could get a signicant
cost reduction, mostly by energy sharing. In Ref. [16], a power sharing model was proposed, to distribute locally generated energy
inside an energy community and to increase social. It is stated that the application of this model can be useful for energy communities
that have renewable energy sources owned by the community’s members. However, the model is based on the ability of each member
to self-consume the electricity generated by the local sources and avoid the exchange of power with others. The impacts of an energy
sharing management framework on distribution grids are studied in Ref. [17] by considering heterogeneous community members in
terms of energy consumption. The proposed battery control algorithm, used to increase the consumption of energy generated locally,
considers the constraints of the low voltage distribution grid and can be implemented on a larger scale. Furthermore, the work reported
in Ref. [18] presented an energy sharing model for zero-carbon communities using a Stackelberg game approach. The model incor-
porated reward and punishment mechanisms to promote energy conservation and emissions reduction. It establishes a framework
involving a zero-carbon community operator and multiple prosumers with energy storage systems scheduled to minimize carbon
emissions. Ceglia et al. compared the benets of a REC with alternative single end users’ congurations and the current reliance on the
power grid [19]. By equipping users with a photovoltaic system, the authors demonstrate increased energy self-consumption and
self-sufciency, and signicant reductions in primary energy demand, carbon dioxide emissions, and operation costs. The research
conducted in Ref. [20] proposed a two-stage stochastic sharing model for communities aiming to maximize the efcient utilization of
PV systems. The model addressed PV power uncertainties through regulation methods, such as demand response and energy storage
charging/discharging scheduling. By employing a data-driven uncertainty set, the model optimized the social cost of PV prosumers and
the community energy storage, resulting in decreased energy costs according to numerical simulations. The concept of peer-to-peer
(P2P) trading platform, in which a member can buy or sell energy to other members, has also been studied in the context of energy
sharing. Such platforms can lead to energy cost savings and local increase of PV electricity usage [14], provide local balance between
supply and demand [21], can be used for both electricity and heating [22], and increase the penetration of electric vehicles in low
voltage grids [23].
Despite the growing interest in energy communities’ management frameworks across Europe, to the best of our knowledge, there is
a research gap regarding the assessment of possible sharing strategies based on energy sharing coefcients dened according to
existing national legislation, as it is the case of Decree-Law 15/2022 [7] in Portugal, and their consequential impacts on energy bills.
Therefore, to address this gap, the present work aims to assess the benets provided to REC members by different energy sharing
coefcients proposed by the Portuguese legislation. This assessment is made by simulating a REC over a year using real consumption
and meteorological data while applying different sharing coefcients (xed and variable). Furthermore, different energy management
schemes are considered for REC members, namely: i) individual operation of the buildings with no association as a community; ii) REC
operation by performing collective self-consumption; iii) and REC operation while also considering individual self-consumption.
1.3. Structure of the document
The paper is organized as follows. Section 2 brings the methodology used in this study, describing the data used to simulate the
operation of the considered buildings, and the algorithms applied to the management of generation units and to the operation of the
H. Queiroz et al.

Heliyon 9 (2023) e20599
4
considered REC. Section 3 presents the collected results and the respective analysis. Lastly, Section 4 addresses the conclusions and
future research directions.
2. Methodology
The assessment of the possible sharing processes described in the Portuguese regulation is made by the simulation of an energy
management system for, both, individual buildings, and an association of buildings as a REC. The implementation of a collective
energy management system is therefore needed according to the Portuguese legal framework [7,8] to compute the electricity con-
sumption, generation, sharing and export, and the associated costs and revenues, if applicable. The buildings considered for the case
study belong to ´
Evora municipality, a historical city in the centre of Portugal, as described below:
•Building 1 (Bld#1) – a multipurpose pavilion, used for concerts, sports tournaments, and other events;
•Building 2 (Bld#2) – the City Hall building;
•Building 3 (Bld#3) – a public market;
•Building 4 (Bld#4) – a public school.
2.1. Electricity demand prole
To assess the impacts of the association of buildings as a REC, real data regarding the buildings’ electricity consumption were
collected. The time frame of data gathering is the whole year of 2019, and the data resolution is 15 min since the data were collected
directly from the Distribution System Operator (DSO) via a specic request by the buildings’ owner and the following acquirement on
the DSO database, which is a common procedure for buildings equipped with smart meters in Portugal. The buildings selected for the
case study, which are part of one of the Lighthouse Pilots of the H2020 POCITYF project [24], have different electricity consumption
proles due to the following factors: size, usage type, and year of construction. For instance, since Bld#4 is a school, there is a sig-
nicant decrease of electricity consumption from the end of July to the second week of September, corresponding to the summer
holiday period for students and most school workers. On a daily perspective, Bld#1 is used for concerts, and other events, and has a
demand peak during the night (around 08:00 p.m.), while in Bld#3, as a market, the demand peak occurs early in the morning. Despite
the holiday period mentioned for Bld#4, its demand peak occurs around noon, almost at the same time in which it occurs for Bld#2.
These particularities can affect PV self-consumption and the impact on the performance of the considered REC. Fig. 1 shows the
normalised daily average electricity demand prole for each building, and Table 1 shows their respective contracted power (i.e., the
maximum power capacity contracted for each building).
2.2. Electricity generation prole
PV systems are used to generate electricity on-site in each building. The electricity generated by each equipment is calculated based
on the generation model described by Ref. [25]. For a PV system, the instant power output (PG(n)) is described by Equation (1), where
N is the number of modules, A is the useful area of one of each PV module, G(n)is the instantaneous global solar irradiance on the
horizontal plane, at the considered local,
η
c(n)is the instantaneous PV module’s efciency, given by Equation (2), and
η
i is the ef-
ciency considering balance-of-system losses from, for instance, power electronics equipment (e.g., DC/AC inverter), cables, or soiling
in the modules (i.e., losses not related with
η
c). For the purpose of this study, a constant value of 0.8 is considered for
η
i as in Ref. [25].
Since this study considers a single year of operation, the yearly degradation of PV modules’ production, often assumed to be 0.5–1.0%
Fig. 1. Normalised buildings’ daily average demand prole.
H. Queiroz et al.

Heliyon 9 (2023) e20599
5
per year [3], was not considered.
PG(n) = N×A×G(n) ×
η
c(n) ×
η
i(1)
η
c(n) =
η
STC ×{1+
μ
×[θa(n) − Tc,STC +G(n) × (Tc,NOCT −θa,NOCT
GNOCT )× (1−
η
STC)]} (2)
In Equation (2),
η
STC is the PV module’s efciency at Standard Test Conditions (STC),
μ
is the temperature coefcient, θa(n)is the
ambient temperature, in degree Celsius, Tc,STC is the PV cell temperature at STC, Tc,NOCT is the cell temperature at Nominal Operating
Cell Temperature (NOCT), θa,NOCT is the ambient temperature at NOCT and GNOCT is the global irradiance at NOCT. The module pa-
rameters used in this study are presented in Table 2. The number of modules for each building (N), used to calculate the PV system
output power in Equation (1), results from a survey conducted during the POCITYF project [24], and are presented in Table 3, together
with the total installed power and annual generation considering the meteorological data described below in this section. PV modules
are assumed to be installed with azimuth ant tilt angles of 0◦and 38◦, respectively, to maximize annual generation, and achieve a
specic production of approximately 1,350 kWh/kWp. The considered buildings are located inside an architecturally sensitive area (i.
e., ´
Evora city centre) and discussions with local authorities are still ongoing to make sure that the nal PV system designs will not
impact the existing cultural heritage. Therefore, the nal PV systems design, including installed capacity, might present some changes
when compared to the values shown in Table 3. However, eventual future modications are believed to not impact the main ndings of
this study as these will not change the considered types of energy sharing coefcients.
Regarding meteorological data, the solar irradiance was collected throughout 2019, with 1-min resolution, by a meteorological
station installed locally. This station is equipped with a two-axis fully automatic sun tracker SOLYS2, with two CMP11 pyranometers to
monitor global horizontal and diffuse solar irradiance, and one CHP1 pyrheliometer to measure direct solar irradiance. The method
described in Ref. [26] was used to convert solar irradiance values from the horizontal plan to the considered inclination of the PV
modules. Due to lack of real ambient temperature data for the considered period (2019), a historic dataset with 1-h resolution was
acquired from the Photovoltaic Geographical Information System (PVGIS), which is an online tool and database developed by the Joint
Research Centre (JRC) of the European Commission [27].
Both irradiance and ambient temperature data were treated to present a 15-min resolution in accordance with the electricity
consumption data referred in Section 2.1. Fig. 2 presents the considered meteorological data with 15-min resolution for 2019 (ambient
temperature in Fig. 2a and solar irradiance in Fig. 2b), together with the respective daily average values.
The installation of storage devices is not considered in this case study. Therefore, surplus electricity is entirely exported to the low
voltage distribution grid. Despite the operation framework chosen for the buildings’ generation units, Fig. 3 shows the algorithm for
the PV electricity management, in which P(n) represents the PV system production at a given instant n, D(n) refers to a specic
building’s demand at the same instant, while ND(n) and M(n) are used to model the building demand after the PV self-consumption
and the exported power after considering the PV self-consumption, respectively. Depending on the chosen operation, the amount of
surplus will change (e.g., if the REC manager allows the members to self-consume part of the locally generated electricity, there will be
less surplus to be shared among them). This rule-based algorithm considers only the local conditions to determine the share of
electricity to be self-consumed or to be exported to the electricity grid. The variable M(n) considers the power readings at the
buildings’ energy meter level. Positive values of M(n) indicate that the PV system was able to provide all the instant electricity needs of
the building and at the same time there is surplus electricity to be exported to the grid. In this case, the building needs after the self-
consumption process, indicated by the variable ND(n), are equal to zero. On the other hand, if M(n) is equal to zero, the PV system was
Table 1
Buildings’ contracted power.
Bld Contracted power [kVA]
1 63.00
2 135.00
3 41.41
4 43.00
Table 2
PV module parameters.
Parameter Value Unit
Rated power 300 W
A 1.63 m
2
η
i 0.8 –
η
STC 0.184 –
μ
−0.0038 ◦C
−1
Tc,STC 25 ◦C
Tc,NOCT 45 ◦C
θa,NOCT 20 ◦C
GNOCT 800 W.m
−2
H. Queiroz et al.

Heliyon 9 (2023) e20599
6
not able to generate all the needs for the building, and ND(n) indicates that need in terms of demand, as the difference between the
original demand and the local production P(n), which is satised by energy import from the grid.
2.3. Renewable Energy Community
According to Ref. [8], a Collective Self-consumption Manager Entity (CSME) must be chosen to form a REC. This entity is
responsible for the commercial relationships between the REC and the remaining stakeholders, such as the DSO and local markets. The
main responsibility of the CSME refers to the electricity sharing among the community members. The CSME is also responsible to
inform the DSO about the energy sharing process chosen by the REC. This sharing process is made by applying energy sharing
Table 3
Buildings’ installed power and PV generation.
Building Number of modules Installed PV power Annual generation
Bld#1 24 modules 7.2 kWp 9.74 MWh
Bld#2 285 modules 85.5 kWp 115.67 MWh
Bld#3 52 modules 15.6 kWp 21.2 MWh
Bld#4 135 modules 40.5 kWp 54.9 MWh
Fig. 2. Meteorological data: (a) Ambient temperature; (b) Solar irradiance.
Fig. 3. PV electricity management algorithm.
H. Queiroz et al.

Heliyon 9 (2023) e20599
7
coefcients, which are related to the percentage of the locally generated energy that is due to each member. The sharing coefcients
can be freely assigned, or calculated, by the rules agreed upon by all REC members [7,8]. The current possibilities, according to the
most recent Portuguese legislation, are the following [8]:
•Fixed coefcients for each member during a considered time frame, being the values freely decided by the REC members (they can
be equal for each member or different among them).
•Variable coefcients computed by any valid criteria (e.g., measured consumption) as long as agreed by all REC members and using
the time interval dened on the Portuguese regulation, namely 15 min [28].
While xed coefcients are easier to apply, the variable ones offer the ability to adapt the sharing process according to specic
needs of the REC under consideration, such as situations in which some of the buildings do not need to import electricity due to its prior
self-consumption or to avoid sharing excessive electricity to a given building, compromising the operation of the REC as a whole. A
general formulation for the calculation of this sharing coefcient (ShCoeff) during a given timestep n is described by Equation (3),
where A, B, and C represent the considered parameters for the sharing coefcient calculation (e.g., energy demand or generation),
while
α
, β, and γ are the respective ponderation factors.
ShCoeff(n) = 100 ×⎡
⎢
⎢
⎣
α
×⎛
⎜
⎜
⎝
Ai(n)
∑
N BLD
i=1
Ai(n)
⎞
⎟
⎟
⎠
+β×⎛
⎜
⎜
⎝
Bi(n)
∑
N BLD
i=1
Bi(n)
⎞
⎟
⎟
⎠
+γ×⎛
⎜
⎜
⎝
Ci(n)
∑
N BLD
i=1
Ci(n)
⎞
⎟
⎟
⎠
+…⎤
⎥
⎥
⎦
(3)
This study only considers the REC member electricity needs, regardless the permission to individual self-consumption. However,
other parameters can be dened to calculate ShCoeff, such as the share of each member to the RES generated energy, the contribution
of each member to the CAPEX/OPEX of equipment to allow the REC operation etc. Also, subjective parameters can be considered, such
as, the level of energy poverty of each member.
Fig. 4 shows the REC conguration used in this work, which will considered both xed and variable energy sharing coefcients (see
Section 3.1 for more information on the considered scenarios). The renewable energy project mentioned by Decree-Law 15/2022 [7] is
satised in this study by the PV systems installed inside the limits of each member of the considered REC, connected to the distribution
grid approximately at the energy meter location. This REC is also dened under the scope of the POCITYF project [24].
It is important to highlight that the shared electricity is part of the total electricity generated within the REC, being assigned to a
given member after the application of the energy sharing coefcient. However, the energy sharing process among the members is made
by the DSO after the billing period, by analysing the respective energy meters’ data and proceeding to a nancial operation to share the
electricity, based on the coefcients informed by the CSME, due to the physical impossibility to drive the electricity instantly to
whichever building that has an energy share assigned. Regarding the variable sharing coefcients, they are only informed after the
billing period, being calculated at each 15-min interval of energy generation and consumption. This operation is made by the addition
or subtraction of the measured power, considering the available power to be shared at each instant. After the sharing process, the
electricity for billing is calculated considering the modied load diagram [8]. However, despite the sharing process being a nancial
Fig. 4. REC conguration.
H. Queiroz et al.

Heliyon 9 (2023) e20599
8
operation conducted after the real consumption and generation of electricity, there may be an actual alleviation of the power grid
during the operation of the local PV systems, either when the buildings are self-consuming, or when electricity surplus exports occur.
This procedure is in line with the ambitions of the European Commission regarding the increase of the usage of renewable energy
sources close to the consumption site [5]. The designed algorithm for the electricity sharing for the REC member is shown on Fig. 5,
being the input and output parameters described in Table 4.
It is important to note that in some cases REC members can self-consume part of the locally generated electricity, which can lead to
possible net-zero power balances. This situation may allow a greater share of electricity for the remaining buildings which are not
capable to achieve a net-zero power balance during a specic time-step. Despite the same algorithm being used on both REC con-
gurations (i.e., REC operation with and without individual self-consumption), the input values are different. BDi(n)refers to the
building demand after the individual self-consumption procedure takes place if this procedure is permitted. When only the collective
self-consumption is allowed, the consumption needs are represented by the sum of all the electric operating loads and the power
available to be shared is the total power production of the locally installed PV systems. Similarly, BPi(n)refers to the power available
for sharing at each time instant. When the prior individual self-consumption is allowed by the REC members, BPi(n)represents the
power surplus after this process. If the REC operation is based on collective self-consumption, BPi(n)represents the power locally
produced by all members. This framework leads to a difference on the amount of energy available for sharing among the REC members,
which depends on how they decide to manage the locally generated electricity. If the decision is to proceed to an individual self-
consumption before the sharing process, the available amount of electricity to be shared is lower, since part of the generated elec-
tricity is consumed by the building users.
Fig. 5. REC energy sharing algorithm.
H. Queiroz et al.

Heliyon 9 (2023) e20599
9
3. Results and analysis
The REC performance for the different scenarios is compared considering the annual cost with electricity import (C
imp
), the amount
of electricity imported (E
imp
), the total of electricity shared among the members (E
sh
), and the amount of electricity exported to the
distribution grid (E
exp
). All simulations were carried out using MATLAB™ software tool.
3.1. Considered scenarios
This study considers the following scenarios: i) baseline operation, representing the current individual energy consumption of the
buildings (no PV systems installed); ii) individual operation with PV systems installed at each building according to the information
provided in Section 2.2; iii) REC operation with buildings able to perform individual self-consumption and sharing only the surplus of
the locally generated energy; and iv) REC operation with buildings performing collective self-consumption of the total electricity
generated by the considered PV systems.
Additionally, to assess the impacts introduced by different energy sharing options on the performance of the referred REC, this
study considers three types of energy sharing coefcients in Scenarios 3 and 4, which covered the possibilities offered under the
Portuguese legislation described in Section 2.3. The rst one refers to a 25% xed and equal coefcient (FEC), so all members receive
the same amount of energy surplus. The second type refers to a xed and proportional coefcient (FPC). In this case, the amount of
energy surplus received by each member is proportional to its relative annual energy consumption (comparing to the annual con-
sumption of the four buildings). The last one refers to a variable sharing coefcient (VSC), dened at each 15-min time-step n, ac-
cording to the relative energy consumption of each member, as described by Equation (4), where VSC
i
(n) refers to the variable sharing
coefcient attributed to building i, while BD
i
(n) represents the individual power demand of each building and N BLD is the number of
REC members. Equation (4) is therefore a particular case of the sharing coefcient described by Equation (3), with a single ponderation
factor to be considered. Table 5 presents a summary of the considered scenarios.
VSCi(n) = BDi(n)
∑
N BLD
j=1
BDj(n)
(4)
3.2. Individual operation without PV systems
In this baseline scenario there is no local electricity generation, thus no sharing is available. For the calculation of the associated
costs, the data gathered, and rules described in Ref. [28] are used. All the considered buildings have the following characteristics:
Table 4
Inputs and outputs for the REC sharing algorithm.
Parameter Description Unit
i Reference to each REC member –
com Reference to the REC as a whole
n Any given time-step considered for the operation –
N_BLD Total of REC members –
Inputs
BD Building demand in REC operation kW
BP Building production available for sharing kW
SC Energy sharing coefcient %
Mid-Process
ESC REC Export sharing coefcient %
Outputs
S Power to share to each REC member kW
NBD Building demand after sharing process kW
E Power exported attributed to each REC member kW
Table 5
Scenarios summary.
3.2) Scenario 1: Individual operation (without PV systems). 3.3) Scenario 2: Individual PV Self-consumption (without REC
association).
3.4) Scenario 3: REC operation with individual PV self-consumption and sharing of energy
surplus.
3.5) Scenario 4: REC operation with collective PV self-consumption.
Energy Sharing Coefcients for REC scenarios:
•Fixed and equal (FEC)
•Fixed and proportional to the annual building consumption (FPC)
•Variable for each measured 15-min. time-step, proportional to each building consumption (VSC)
H. Queiroz et al.

Heliyon 9 (2023) e20599
10
•The voltage level for electricity supply at the energy meter level, considered for the tariff application, is 400 V (line-to-line voltage);
•The daily billing cycle is applied, which means that the instants of the application of each tariff are constant during the entire week
(there is no difference between the tariffs on weekdays and weekends);
•There are four different tariffs to be applied during each daily cycle, namely: peak, half-peak, normal off-peak and super off-peak.
Table 6 shows the periods of each tariff, and the corresponding values.
Additionally, for the calculation of the electricity costs, a xed tariff (24.64
€
/month) was considered, and a contracted power tariff
(0.69
€
/kW.month) applied, together with a value for the power demand in peak hours, calculated as the average of the demand power
during peak hours (15.64
€
/kW.month). These values were obtained by the analysis of an electricity bill from one of the buildings and
applied to all of them. The considered metrics for the assessment of Scenario 1 are presented in Table 7.
3.3. Individual operation with PV systems
In this scenario, there is energy export (E
exp
) due to the mismatch between consumption needs and energy locally generated by the
PV systems described in Section 2.2. The power export is calculated by using the algorithm showed in Fig. 3. Table 8 shows the in-
dicators for this scenario. The amount of energy self-consumed by each building (E
sc
) is also presented. There is a natural decreasing on
the imported energy and associated costs, due to the energy generated by the PV systems, which is partially used to full the building
user’s needs. The referred mismatch results in the total export of 34% of all electricity generated by the local sources. The remaining
66% refers to the generation self-consumed by the buildings.
3.4. – REC operation with sharing of individual generation surplus
This conguration implies that buildings prioritize individual self-consumption of local generation, being the eventual surplus
shared among the members that, on a given time-step, were not able to full their electricity needs with the respective PV systems.
Therefore, the amount of available energy to the REC sharing must be calculated at every time-step, by subtracting, from the generated
energy, the energy self-consumed by the building.
As aforementioned, three different sharing coefcients are considered for each one of the REC congurations. Table 9 shows the
calculated values for the considered xed sharing coefcients (values for the variable sharing coefcient (VSC) are not depicted here
given that a different value is dened for each one of the 35,040 (96 ×365) time-steps).
The operation results are shown in Table 10, in which, for each indicator and energy sharing coefcient for a specic building,
values marked in a dark grey indicate the worst performance, light grey indicate the best performance and the remaining grey values
indicate a performance between the previous ones. Note that the amount of energy self-consumed (E
sc
) in this scenario is the same as in
Scenario 2, due to the individual self-consumption occurring prior to the sharing process. The values of self-consumed energy can be
seen in Table 8.
3.5. – REC operation with collective self-consumption
The main characteristic of this scenario is that the PV systems are not used for individual self-consumption before the sharing
process takes place, resulting in more energy to be shared among REC members. The remaining process regarding the energy sharing is
the same as in the previous scenario, including the value of both xed coefcients. Table 11 summarizes the performance indicators
and the same colour pattern of Table 10 is applied.
3.6. Discussion
The collected results show that the operation as a REC (Scenarios 3 and 4) conducts to lower energy costs and less imported energy,
when compared to their individual operation (Scenario 2). This results from energy sharing among REC members, when considering
the possibility of individual self-consumption (Scenarios 3) or collective self-consumption (Scenarios 4), which also leads to less energy
exported to the distribution grid. Scenario 1, where no PV systems are available, is the one with the worst performance in terms of costs
Table 6
Daily cycle tariff intervals and values.
Winter legal hours Summer legal hours
Peak 09h00 – 10h30 Peak 10h30 – 13h00
(0.21
€
/kWh) 18h00 – 20h30 (0.21
€
/kWh) 19h30 – 21h00
Half-peak 08h00 – 09h00 Half-peak 08h00 – 10h30
(0.13
€
/kWh) 10h30 – 18h00 (0.13
€
/kWh) 13h00 – 19h30
20h30 – 22h00 21h00 – 22h00
Off-peak 06h00 – 08h00 Off-peak 06h00 – 08h00
(0.09
€
/kWh) 22h00 – 02h00 (0.09
€
/kWh) 22h00 – 02h00
Super off-peak 02h00 – 06h00 Super off-peak 02h00 – 06h00
(0.08
€
/kWh) (0.08
€
/kWh)
H. Queiroz et al.

Heliyon 9 (2023) e20599
11
and imported energy. To support this analysis, Fig. 6 shows the daily average demand prole for all buildings, scenarios, and sharing
coefcients (xed and equal sharing coefcients in Fig. 6a, xed and proportional sharing coefcients in Fig. 6b, and variable sharing
coefcients in Fig. 6c). In this gure, Scenario 2 is added for the sake of comparation with the REC operation (since no energy sharing
coefcients are applied), while Scenario 1 is not considered given the inexistence of local generation.
Regarding the energy sharing coefcients, one can conclude from Tables 10 and 11 that selecting the best option depends on the
considered scenario and demand prole of each building. However, it is observed that the VSC leads to a better REC performance, from
a collective point of view, with lower energy costs and less energy imported. Nevertheless, when individual self-consumption is
allowed for the REC members (Scenario 3), there is a lower advantage from choosing the VSC, due to the individual benet already
achieved by the individual self-consumption and less energy needs when the sharing process happens. This leads to an improved
sharing process, specically when the VSC is applied. Additionally, Scenario 4 presents the downside of inhibiting individual self-
consumption from the PV systems installed on each building, which, considering investment costs, can bring challenges to engage
members in selecting a REC operation with collective self-consumption over a REC operation where individual self-consumption is
allowed (Scenario 3). In this case, considering a VSC according to other factors besides the buildings’ demand (e.g., investment costs)
might benet the engagement process (see Equation (3)).
Other important aspect to consider is the difference on the buildings’ energy needs and local generation. Table 12 shows the annual
generation and demand for each building, together with the ratio between these two values, for each building. The weight of Bld#2 on
the sharing process is higher in any REC conguration due to its larger energy needs and therefore it benets the most from the VSC, as
can it be seen in Fig. 7, which presents the individual performance of each building per scenario (Bld#1 in Fig. 7a, Bld#2 in Fig. 7b,
Bld#3 in Fig. 7c, and Bld#4 in Fig. 7d). Bld#1 and Bld#3, with the lowest energy demand, show the best results when the FEC is applied
as the amount of energy to be shared is not impacted by the energy needs of other members. However, even with this energy demand
unbalance, the variable sharing coefcient can improve the performance of the whole REC comparing with the other possibilities of
energy sharing (see Fig. 6a–c).
It is important to note that the operation framework described here refers only to nancial operations, conducted by the DSO, after
each monthly billing period throughout the year. Regardless the energy sharing framework, in every scenario of REC operation there is
the same amount of energy involved. The main difference is the way the sharing is carried out, which impact the performance in terms
of energy costs.
4. – Conclusions and future work
The study reported in this paper focus the assessment of different energy sharing strategies to be applied in Renewable Energy
Communities (RECs) across Europe based on energy sharing coefcients dened according to existing national legislation, as it is the
case of Decree-Law 15/2022 in Portugal. This involves simulating a REC over a year, utilizing real consumption and meteorological
Table 7
Scenario 1 indicators.
Bld C
imp
[
€
] E
imp
[kWh]
1 8,353 37,904
2 53,902 258,350
3 7,446 33,289
4 8,545 36,392
Total 78,246 365,935
Table 8
Scenario 2 indicators.
Bld C
imp
[
€
] E
imp
[kWh] E
exp
[kWh] E
sc
[kWh]
1 6,685 30,157 1,994 7,747
2 33,673 165,900 23,225 92,450
3 4,643 21,102 9,054 12,187
4 4,149 16,781 35,224 20,154
Total 49,150 233,940 69,497 132,538
Table 9
Fixed sharing coefcients.
Bld FEC (%) FPC (%)
1 25 10.35
2 25 70.58
3 25 9.10
4 25 9.97
H. Queiroz et al.

Heliyon 9 (2023) e20599
12
data, and implementing various sharing coefcients, both xed and variable. Additionally, this study considers two distinct models for
the REC operation to further understand the benets offered to all members. The rst model allows buildings to conduct individual
self-consumption, while only sharing generation surplus. The second model refers to a scenario where only collective self-consumption
is allowed.
The collected results show that the benets offered to each member depend on the considered model for the REC operation and on
the individual energy needs. Fixed energy sharing coefcients tend to be more interesting to buildings with lower energy needs, while
larger consumers take more advantage from the variable energy sharing coefcient. However, it can be concluded that this last co-
efcient leads to the best performance when considering the entire REC for both models of operation. This study also shows that the
benets for buildings operating as a REC are evident when comparing to their individual operation without generation surplus sharing,
with savings reaching 44% for the building with the highest energy demand.
Further research on different case studies, such as mixed-purpose or net-zero energy buildings joining the same REC, or the impact
of energy storage equipment on the community’s performance, is needed. Additionally, new formulations for the variable energy
sharing coefcients should be studied. On a scenario of renewable energy communities’ signicant market growth, further investi-
gation is needed on how to mitigate possible challenges imposed by RECs on distributions grids, such as power congestion or voltage
rising problems, due to the integration local energy generation units.
Author contribution statement
Humberto Queiroz: conceived and designed the experiments; performed the experiments; analysed and interpreted the data;
contributed reagents, materials, analysis tools or data; wrote the paper. ui Amaral Lopes: conceived and designed the experiments;
analysed and interpreted the data; wrote the paper. Jo˜
ao Martins, Luís Fialho: analysed and interpreted the data; contributed reagents,
materials, analysis tools or data; wrote the paper. Filipe Neves Silva: analysed and interpreted the data; wrote the paper. Nuno Bilo:
contributed reagents, materials, analysis tools or data; wrote the paper.
Table 10
Scenario 3 indicators.
H. Queiroz et al.

Heliyon 9 (2023) e20599
13
Table 11
Scenario 4 indicators.
Fig. 6. Normalised REC average load diagram: (a) FEC; (b) FPC; (c) VSC.
H. Queiroz et al.

Heliyon 9 (2023) e20599
14
Data availability statement
The data that has been used is condential.
Declaration of Competing Interest
The authors declare that they have no known competing nancial interests or personal relationships that could have appeared to
inuence the work reported in this paper.
Acknowledgements
This research was partly funded by European Union’s H2020 programme as part of the POCITYF project (a positive energy city
transformation framework), Grant agreement ID 864400 and by the Portuguese “Fundaç˜
ao para a Ciˆ
encia e a Tecnologia” (FCT) in the
context of the Center of Technology and Systems CTS/UNINOVA/FCT/NOVA, reference UIDB/00066/2020.
References
[1] European Commission, The European green deal, Eur. Community 53 (2019) 24, https://doi.org/10.1017/CBO9781107415324.004.
[2] R.A. Lopes, P. Magalh˜
aes, J.P. Gouveia, D. Aelenei, C. Lima, J. Martins, A case study on the impact of nearly Zero-Energy Buildings on distribution transformer
aging, Energy 157 (2018) 669–678, https://doi.org/10.1016/j.energy.2018.05.148.
[3] H. Queiroz, R. Amaral Lopes, J. Martins, Automated energy storage and curtailment system to mitigate distribution transformer aging due to high renewable
energy penetration, Elec. Power Syst. Res. 182 (2020), 106199, https://doi.org/10.1016/j.epsr.2020.106199.
Table 12
Annual generation and demand, together with the ratio between these metrics.
Bld Generation [kWh] Demand [kWh] Generation to demand ratio [%]
1 9,741.03 37,903.75 25.70%
2 115,674.68 258,350.00 44.77%
3 21,240.85 33,289.00 63.81%
4 54,928.56 36,392.25 150.93%
Total 201,585.12 365,935.00 55.09%
Fig. 7. Buildings’ daily average demand prole per scenario: (a) Bld#1; (b) Bld#2; (c) Bld#3; (d) Bld#4.
H. Queiroz et al.

Heliyon 9 (2023) e20599
15
[4] R.A. Lopes, J. Martins, D. Aelenei, C.P. Lima, A cooperative net zero energy community to improve load matching, Renew. Energy 93 (2016) 1–13, https://doi.
org/10.1016/j.renene.2016.02.044.
[5] European Parliament and the Council of the European Union, Directive (Eu) 2018/2001 Of The European Parliament And Of The Council Of 11 December 2018
On The Promotion Of The Use Of Energy From Renewable Sources (Recast), 2018.
[6] European Parliament and the Council of the European Union, Directive (Eu) 2019, 2019. /944 Of The European Parliament And Of The Council Of 5 June 2019
On Common Rules For The Internal Market For Electricity And Amending Directive 2012/27/Eu (Recast).
[7] Presidˆ
encia do Conselho de Ministros, Decreto-Lei n.
o
15/2022 de 14 de janeiro, 2022.
[8] Entidade Reguladora dos Serviços Energ´
eticos, Regulamento n.
o
373/2021 (2021).
[9] M.-H. Kim, D. Kim, J. Heo, D.-W. Lee, Techno-economic analysis of hybrid renewable energy system with solar district heating for net zero energy community,
Energy 187 (2019), 115916, https://doi.org/10.1016/j.energy.2019.115916.
[10] F. Ceglia, E. Marrasso, C. Roselli, M. Sasso, Small renewable energy community: the role of energy and environmental indicators for power grid, Sustainability
13 (2021) 2137, https://doi.org/10.3390/su13042137.
[11] A. Mar, P. Pereira, J. Martins, Energy community exibility solutions to improve users’ wellbeing, Energies 14 (2021), https://doi.org/10.3390/en14123403.
[12] H. Firoozi, H. Khajeh, H. Laaksonen, Optimized operation of local energy community providing frequency restoration reserve, IEEE Access 8 (2020)
180558–180575, https://doi.org/10.1109/ACCESS.2020.3027710.
[13] F. Ceglia, A. Macaluso, E. Marrasso, C. Roselli, L. Vanoli, Energy, environmental, and economic analyses of geothermal polygeneration system using dynamic
simulations, Energies 13 (2020) 4603, https://doi.org/10.3390/en13184603.
[14] N. Liu, X. Yu, C. Wang, C. Li, L. Ma, J. Lei, An energy sharing model with price-based demand response for microgrids of peer-to-peer prosumers, IEEE Trans.
Power Syst. 32 (2017) 3569–3583, https://doi.org/10.1109/TPWRS.2017.2649558.
[15] D. Fioriti, A. Frangioni, D. Poli, Optimal sizing of energy communities with fair revenue sharing and exit clauses: value, role and business model of aggregators
and users, Appl. Energy 299 (2021), 117328, https://doi.org/10.1016/j.apenergy.2021.117328.
[16] L. Martirano, S. Rotondo, M. Kermani, F. Massarella, R. Gravina, Power sharing model for energy communities of buildings, IEEE Trans. Ind. Appl. 57 (2021)
170–178, https://doi.org/10.1109/TIA.2020.3036015.
[17] S. Norbu, B. Couraud, V. Robu, M. Andoni, D. Flynn, Modeling economic sharing of Joint assets in community energy projects under LV network constraints,
IEEE Access 9 (2021) 112019–112042, https://doi.org/10.1109/ACCESS.2021.3103480.
[18] H. Gao, W. Cai, S. He, C. Liu, J. Liu, Stackelberg game based energy sharing for zero-carbon community considering reward and punishment of carbon emission,
Energy 277 (2023), 127629, https://doi.org/10.1016/j.energy.2023.127629.
[19] F. Ceglia, E. Marrasso, C. Martone, G. Pallotta, C. Roselli, M. Sasso, Towards the decarbonization of industrial districts through renewable energy communities:
techno-economic feasibility of an Italian case study, Energies 16 (2023) 2722, https://doi.org/10.3390/en16062722.
[20] B. Gu, C. Mao, D. Wang, B. Liu, H. Fan, R. Fang, Z. Sang, A data-driven stochastic energy sharing optimization and implementation for community energy
storage and PV prosumers, Sustain. Energy, Grids Networks 34 (2023), 101051, https://doi.org/10.1016/j.segan.2023.101051.
[21] J. Guerrero, A.C. Chapman, G. Verbic, Decentralized P2P energy trading under network constraints in a low-voltage network, IEEE Trans. Smart Grid 10 (2019)
5163–5173, https://doi.org/10.1109/TSG.2018.2878445.
[22] R. Jing, M.N. Xie, F.X. Wang, L.X. Chen, Fair P2P energy trading between residential and commercial multi-energy systems enabling integrated demand-side
management, Appl. Energy 262 (2020), 114551, https://doi.org/10.1016/j.apenergy.2020.114551.
[23] T.D. Hutty, A. Pena-Bello, S. Dong, D. Parra, R. Rothman, S. Brown, Peer-to-peer electricity trading as an enabler of increased PV and EV ownership, Energy
Convers. Manag. 245 (2021), 114634, https://doi.org/10.1016/j.enconman.2021.114634.
[24] POCITYF Project, About POCITYF, 2022. https://pocityf.eu/. (Accessed 4 February 2022).
[25] J. Wid´
en, E. W¨
ackelgård, J. Paatero, P. Lund, Impacts of distributed photovoltaics on network voltages: stochastic simulations of three Swedish low-voltage
distribution grids, Elec. Power Syst. Res. 80 (2010) 1562–1571, https://doi.org/10.1016/j.epsr.2010.07.007.
[26] J.A. Dufe, W.A. Beckman, Solar Engineering of Thermal Processes, fourth ed., John Wiley & Sons, Inc., Hoboken, NJ, USA, 2013 https://doi.org/10.1002/
9781118671603.
[27] E.C. Joint Research Center (JRC), Photovoltaic Geographical Information System (PVGIS), 2022. https://joint-research-centre.ec.europa.eu/photovoltaic-
geographical-information-system-pvgis_en. (Accessed 11 July 2023).
[28] Entidade Reguladora dos Serviços Energ´
eticos, Regulamento Tarif´
ario - Setor El´
etrico, 2020. https://www.erse.pt/media/bhnpuida/articulado-rt-se_
consolidado.pdf.
H. Queiroz et al.