Energy Policy 182 (2023) 113745

Contents lists available at ScienceDirect

Energy Policy
journal homepage: www.elsevier.com/locate/enpol

Prosumers integration in aggregated demand response systems

Carlos Cruz a ,∗, Tarek Alskaif b , Esther Palomar a , Ignacio Bravo a
a
Department of Electronics, University of Alcala, Alcala de Henares, 28871, Madrid, Spain
b
Information Technology Group, Wageningen University and Research (WUR), Wageningen, 6706 KN, Netherlands

ARTICLE INFO ABSTRACT

Keywords: In recent years, the increasing popularity of renewable energy and energy-efficient technologies is creating a
Cooperative demand response new movement towards more sustainable communities. Understanding energy consumption is important for
Energy communities the optimisation of resources and the implementation of ecological trends. This paper integrates electricity
Prosumers
consumers into a cooperative framework for planning sustainable smart communities through aggregators,
Renewable energy resources
which reallocates consumers’ demand according to available renewable energy supply collected from con-
Microgeneration
Aggregators
sumers and service providers. The aggregated demand response also includes consumers involved in energy
production activities through microgeneration capabilities. A characterisation study of the different types of
demand preferences is performed by defining scenarios of communities and consumers’ behaviours, which
are validated through a reputation factor. The results show that the system adequately manages demand
reallocation following the preferences and contribution of consumers and/or prosumers. Besides, this research
analyses the current energy policy concerning demand flexibility, demand aggregation and microgeneration
capacity, and their regulations in Spain. Finally, microgeneration acceptance, the role of aggregators and
prosumers in the scheduling process is also investigated through a series of surveys.

1. Introduction (70%) to the international increase in gas price and CO2 emission
rights.
Recent Intergovernmental Panel on Climate Change (IPCC) reports The European Union (EU) is taking actions to adapt to the impacts
drawn attention about the state of scientific, technical knowledge on of climate change with initiatives under the Green Deal. For instance,
climate change and the options for mitigating its impact. In this context, the EU has set energy policy objectives as the Fit for 55 package (Silviu,
cities, smart communities and residential energy demands offer signifi- 2022) based on climate or energy policies to reduce greenhouse gas
cant opportunities for sustainable energy transition and CO2 emissions emissions by at least 55% until 2030. Therefore, Brussels is increasing
reductions. The customer participation in the energy market and the pressure on emissions despite their impact on electricity prices. Stable
integration of Information and Communication Technology (ICT) in energy prices will only come with an increase in renewable generation
smart grids are deemed to become an integral part of residential energy and a less dependence of fossil fuels. In this regard, the Spanish
systems evolution (Camarinha-Matos, 2016). framework for energy and climate sets an objective to reach a national
Three global challenges affect the residential sector: 40% of CO2 climate neutrality and 97% renewable energy in the total energy mix
emissions come from the residential field (Eds, 2022); more than 30% by 2050. Beyond the electricity sector, the Spanish government also
of needed energy in buildings is wasted (EPA, 2021); and approxi- plan to expand self-consumption of renewables and distributed gener-
mately 90% of our time is spent indoors (Hallen, 2021). In addition, ation, as well as promote and encourage the use of renewables in the
the increase in global demand has grown by around 20% over the last residential sector.
20 years and this trend will increase at an annual rate of approximately Smart community is gaining importance as a promising solution to
1.8% over the period 2020–2030 (Brugger et al., 2021). Furthermore, the challenge of sustainable energy communities providing a minimal
the current situation of rising prices on the energy market is taking impact on environment and citizen lifestyles (Bibri and Krogstie, 2017).
place in a context of health crisis caused by COVID-19, which had a Consumers behaviour focused by patterns may be the most preeminent
huge impact on the economic activity. The high electricity expenditure city asset in smart and sustainable power grids. For instance, Kramers
leads to higher prices, and consequently, means regressive effects on et al. (2014) explore opportunities for the reduction of energy consump-
households (Ine, 2021). The rising electricity prices is mainly due tion in cities through an intelligent aggregation of anticipated demand

∗ Corresponding author.
E-mail addresses: [email protected] (C. Cruz), [email protected] (T. Alskaif), [email protected] (E. Palomar), [email protected] (I. Bravo).

https://doi.org/10.1016/j.enpol.2023.113745
Received 11 September 2022; Received in revised form 13 April 2023; Accepted 29 July 2023
Available online 9 August 2023
0301-4215/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
C. Cruz et al. Energy Policy 182 (2023) 113745

this regard, the aggregated DR optimises the time intervals of flexible

Nomenclature
smart appliances and other household devices operation preferences,
𝑓 Fixed demand maximising the consumption of those assets from available renewable
𝑣 Variable demand energy supply without due consideration to microgeration capabilites.
 Consumer number Other studies propose community-based local energy markets that aim
to maximise self-sufficiency within a community by encouraging de-
𝐷𝑅 Demand Response
mand side management and local energy exchange (i.e., photovoltaics
𝐷𝑅𝐶 Demand Reallocation Capacity
systems (PV)), which consequently minimises energy exchange with
𝐹 Demand Function
the wholesale market (Crespo-Vazquez et al., 2020). A cooperative
𝐺𝑃 𝑉 Photovoltaic generation DR management for consumers and prosumers together is a solid
𝐺𝑅𝑊 24-h supply from renewables mechanism for future developments in this regard. The importance of
𝐼𝐶𝑇 Information and Communication Technolo- understanding DR schemes will enable incorporating the behavioural
gies factors of consumers and/or prosumers and how these impact the
𝐼𝑜𝑇 Internet Of Things energy aggregation service’s performance.
𝑅𝐶𝑇 Ratio of the Computation Time
𝑆𝑃 𝑉 Storage Photovoltaic system 1.2. Prosumers
𝑡𝑏𝑒𝑔,𝑖𝑗 Earliest start time appliance
𝑡𝑒𝑛𝑑,𝑖𝑗 Latest final time appliance PV or any electric installations including shared batteries between
𝑡𝑠𝑐ℎ𝑒𝑑,𝑖𝑗 Scheduled start time of appliance neighbours are a good example of smart efficient energy commu-
nities as stated in Gallego-Castillo et al. (2021). Furthermore, the
profitability of PV self-consumption installations is now positive in
several countries, including Spain under the current Spanish regulation
patterns of multiple consumers. New technologies and applications are for microgeneration deployment systems (López Prol and Steininger,
providing opportunities to increase energy efficiency and peak demand 2020). The benefits of self-consumption are mainly reached by sharing
reduction (Silva et al., 2018). To this regard a framework for smart and solar 𝑃 𝑉 energy among the consumers participating in DR (Faria et al.,
sustainable development can help to reduce cost and carbon impact 2019). As an example, Conchado et al. (2016) present an integrated
in residential buildings. Hence, it becomes extremely important to assessment of a potential DR programme in Spanish households, includ-
both intelligently control residential energy consumption, in order to ing both supply and demand side considerations. Kashanizadeh et al.
reduce greenhouse gases and safeguard energy resources (Haidar et al., (2022) analyse the reliability rates and the share of battery energy
2018), and to incentivise consumers to be more active and empowers storage system as a local decentralised generation assets. However, the
them in having more control over their electricity consumption through benefits are quite low compared to the costs, and most of them come
Demand Response (DR) services. Consumers adjust their consumption from the system generation. Most of the research works take advantage
by DR models, which have emerged as an instrument to relieve supply of least-cost pricing while considering prosumers flexibility behaviour
and demand imbalances of electricity (Yamaguchi et al., 2020). and its cooperation potential.
The microgeneration acceptance can be expressed through be-
1.1. Cooperative demand response haviours and investments. Sauter and Watson (2007) and Souza et al.
(2018) show case studies and results that will open up new opportu-
End-users demand flexibility is an important factor for implement- nities for the residential electricity market. Effective deployment will
ing DR programmes. Several studies have analysed the impact of end- partially depend on public attitudes and acceptance at both a commu-
users participation on the success of a DR programme or proposed nity and household levels (Ambrosio-Albala et al., 2020). People are
solutions to increase their participation (Parrish et al., 2020). For more likely to accept energy storage facilities in their neighbourhood
instance, Jovanovic et al. (2016) take into account the preferences of establishing the aggregation or market facilitator figures (Koecklin
households participation, which is modelled according to their ‘‘flexi- et al., 2021). Users may weigh the expected complexity and effort
bility’’. The results show significant savings in electricity costs. Singh against the expected DR benefits in deciding whether to enrol. In terms
and Yassine (2019) demonstrate that appliance associations are a direct of management implementation, the price of the subsidy is crucial
reflection of consumers’ energy use behaviour where an unsupervised and community advertising can also enhance the electricity-saving
data mining process is applied to smart metre energy consumption. effect (Wang et al., 2020). The aggregation is capable of exploiting
However, potential consumers flexibility has been systematically ig- the flexibility of prosumers’ appliances and reducing market net costs
nored due to the highly different behaviours among them (Guo et al., if compared with inflexible strategies (Iria et al., 2018). Current legal
2018). Smart metring systems alone neither automatically recognise framework at the EU level represents a clear opportunity for collective
end-users flexibility nor drive residential customers to use energy in prosumers. Spain has already shifted from a restrictive regulation to
a more sustainable manner. To this regard, gamification-based frame- implementing a legal framework for collectives (Inês et al., 2020).
work was proposed by AlSkaif et al. (2018) to drive behavioural Reputation-based policy can ensure fairness in energy allocation
change of residential customers in relation to energy and increase their in decentralised energy systems and energy communities. It has been
engagement. proposed for microgrids (AlSkaif et al., 2015), blockchain strategies
In a smart energy-efficient buildings, consumers engagement is for interoperation of DR (Liu et al., 2020) or flexible energy mar-
commonly promoted by financial incentives; customers respond to kets (Chukhnina et al., 2021). Reputation factor based systems also be-
incentives programmes (e.g., increases in energy prices) by modify- long to the incentive-based mechanisms in cooperative games (Fiestras-
ing energy usage behaviour as well as investing in energy-efficient Janeiro et al., 2011). Game theory has been widely used in research
technologies and practices (Shareef et al., 2018). On the other hand, works that focus on developing and analysing problems related to
in smart energy community applications, the cooperation of groups renewable energy integration and smart grids. DR systems based on
of consumers towards common goals (e.g., environmentally friendly, reputation are also among the common applications focused on energy
become green) could make a larger impact in sustainable energy man- management in microgrids and energy communities, where there are
agement (Bauwens et al., 2022). For instance, a validation study of types of households with different energy demand profiles. For in-
a cooperative DR framework was proposed in Cruz et al. (2019). To stance, clusters with higher energy generation or cooperation capacity

2
C. Cruz et al. Energy Policy 182 (2023) 113745

can be rewarded appropriately (Guo and Guo, 2015). More specifi- consumers are willing to share their consumption profiles with a third
cally, AlSkaif et al. (2017) deploy an optimisation problem, in which a party aggregation company or with their energy supplier via their smart
demand-side manager jointly schedules the energy consumption of ap- metres. The system consists of a turn-based strategy for a group of con-
pliances and the energy that each household can receive from a shared sumers who are willing to cooperate in achieving the community’s goals
battery storage unit according to their reputation factor. Corporate by sending their preferences to the aggregator. Each consumer has a set
social responsibility initiatives are positively associated with reputation of devices i ∈  labelled . As an application scenario, Fig. 1 presents
and customer satisfaction. To this regard, Islam et al. (2021) propose a visual illustration of the system architecture. It emulates a smart
a mechanism through which corporate social responsibility influences residential energy community staged among different participants:
customer loyalty.
• Consumers and/or prosumers autonomously adapt their energy
1.3. Research gaps and motivation consumption and net generation by means of sharing nearly in
real-time their electricity demand information.
From the literature review, the following limitations can be found • An aggregator is capable of shifting the consumers’ use, and also
in previous research. In relation to the studies carried out in the responsible for the computation and rescheduling of the total
field of DR development for energy communities, it can be easily daily load of the community.
seen that: (1) there is a perceived lack of a comprehensive model
to implement more effective behavioural DR programmes; (2) tradi- 2.1. Appliances demand profile
tional self-consumption policies have a limited potential market for
residential communities without cooperation; The possibility to save The consumer pre-allocates a certain amount of fixed demand and
or share the self-produced energy for a future use may enhance the expected variable consumption for the next 24 h. A large number of ap-
exploitation of renewable energy resources and the reduction of the pliances cannot be completely switched off without being disconnected.
end-users’ energy procurement costs; (3) the value of aggregators and These appliances consume energy 24 h a day and it is referred as
DR aggregation have been extensively studied in recent years. The standby power. Cooling appliances (e.g., fridges) are usually considered
importance of aggregation systems in fostering a large-scale integration as non-shiftable appliances and kettle or dishwashers, for example,
of renewable energy sources is currently overlooked due to the main are considered as dispatchable appliances. We define 𝐷𝑓 and 𝐷𝑣 as
focus on the minimisation of peak load and electricity cost. Therefore, the fixed and variable one-hour energy demand. The total fixed and
there is a need to provide a structure for consumers participation in variable consumption demand for consumer i 𝑁 per day is denoted by
the sustainability of energy communities. To fill these research gaps, 𝐷𝑇 ,𝑖 and given in Eq. (1).
the proposed analysis assumes the management of a cooperative DR
programme which pools the demand of all participants and schedules ∑
23 ∑
𝐷𝑇 ,𝑖 = (𝐷𝑓𝑡 𝑖,𝑎 + 𝐷𝑣𝑖,𝑎
𝑡
) (1)
it according to available renewable energy supply. In order to propose 𝑖𝑗 𝑖𝑗
𝑡=0 𝑎𝑖𝑗 ∈𝑖
energy-saving strategies to achieve a better balance between produc-
tion and consumption, we integrate the renewable energy generation Each household i has a set of dispatchable appliances 𝑖 whose
at the household level. The most important contribution focuses on operation can be scheduled in the next 24 h. Some appliances may be
the categorisation of consumers according to their preferences, and its used more than once a day. Variable energy demand is characterised by
impact on the integration of microgeneration systems. Our proposal its flexibility, as it takes into account the consumers’ preference for an
drastically transforms the distribution networks by turning them into appliance to start in a given time period. For each appliance, there is a
bi-directional, both in the flow of energy and the information, and closed interval that selects a minimum start time and a maximum end
allowing new agents (e.g. aggregator or microgenerators) and energy time labelled by 𝑡𝑏𝑒𝑔 and 𝑡𝑒𝑛𝑑 . The decision variable 𝑡𝑠𝑐ℎ𝑒𝑑,𝑖𝑗 is defined
service companies to compete in order to maximise the renewable as the operation time of the device 𝑗 in household 𝑖 and coincides with
energy consolidation. A deep analysis of consumers behaviour related the start/end time range [𝑡𝑏𝑒𝑔,𝑖𝑗 , 𝑡𝑒𝑛𝑑,𝑖𝑗 ]. The operation time is defined
to reallocation capabilities is performed. The aggregated DR schedul- as the duration of the scheduled operation of the device on the next
ing algorithm enables a feasible demand reallocation in a reasonable day. Hence, the demand must be activated for a time between two
computation time for different consumption patterns by means of a predefined times: ∀𝑖𝑗 ∈ A, 𝑡𝑠𝑐ℎ𝑒𝑑,𝑖𝑗 ≥ 𝑡𝑏𝑒𝑔,𝑖𝑗 . The demand schedule must
reputation system. Regarding the categorisation of consumers’ prefer- also finish before the end time, as: ∀𝑖𝑗 ∈ A, 𝑡𝑠𝑐ℎ𝑒𝑑,𝑖𝑗 ≤ 𝑡𝑒𝑛𝑑,𝑖𝑗 . Consumer 𝑖
ences, we introduce a reputation factor 𝑅 that mitigates the impact of will set the following operation data for each device 𝑗 (i.e., appliance)
user behaviour on the DR framework. 𝑅 is determined for three types 𝑎𝑖𝑗 ∈ . Table 1 shows the most common devices used for different
of behaviour according to demand preferences. Experimentation results types of behaviour defined in Cruz et al. (2021), average consumption,
show patterns where factors such as community demand volume, con- operation time, preferred time and flexibility (Cruz et al., 2019; Torriti,
sumer flexibility and microgeneration, influence the performance of the 2017; Powells et al., 2014; Lo Piano and Smith, 2022).
scheduling system. Finally, we develop a case study with a benchmark
that provides valuable insights into and validation of the proposed 2.2. Renewable power profile
aggregated DR system.
The remainder of this paper is structured as follows. Section 2 Certain number of households have an on-site solar 𝑃 𝑉 system and
presents the methodology, where the demand and generation profiles generate an amount of renewable energy. The degree of microgener-
are constructed, the optimisation problem is introduced and the rep- ation has been selected and sized based on the daily demand profile
utation factor is defined. Section 3 shows the results and discussion. for 𝑃 𝑉 production in household (IEA, 2022). The microgeneration
Finally, Section 4 draws the conclusions and policy implications. capacity is denoted by 𝐺𝑃𝑡 𝑉 = [0, … ., 23]. 𝐺𝑃 𝑉 is an optional variable
for energy demand reallocation at prosumers. The supplier provides
2. Methodology information on the reliable renewable energy sources and the fossil
energy scheduled for the next 24 h using a variable denoted by 𝐺𝑅𝑊 𝑡 =
The proposed system can be considered as a micro network where [0, … ., 23]. 𝐺𝑅𝑊 is essential in optimising a fair renewable energy
a group of consumers, connected to each other, operates with a certain allocation between consumers’ fixed and variable energy demands. It
degree of independence. The methodology assumes that the consumers is available to satisfy the energy resources at the current time slot 𝑡
share their electricity consumption information with the aggregator and a participant decides whether to participate in the cooperation for
for demand reallocation. In practice, this scenario is applicable when the energy provision of the next 24 h. A prosumer has the ability to

3
C. Cruz et al. Energy Policy 182 (2023) 113745

Fig. 1. System design setup of prosumers, an aggregator and energy service providers.

Table 1
List of appliances and their operation data for the different operation scenarios.
Target Level/ Appliance Peak Prefered time per day (in hour) Operation Average consumption Proportion of
temporal likelihood time (kWh) dwellings with
flexibility (No. hour) appliance (%)
Scenario I Scenario II Scenario III Operative Standby
Electric oven High 16 10 3 3 2.13 – 5 0.005 61.6
Lighting Medium/ Microwave High 14 12 4 4 1.25 – 3 0.01 85.9
cooking Sub-hour Refrigerator High 18 10 10 10 1.77 – 4 0.01 96
appliances Kettle High 20 14 6 6 2 – 5 0.005 97.5
Hob High 18 12 2 2 1 – 3 0.01 46.3
Electrical-based Low/ Tumble dryer Middle 18 12 2 2 2.60 – 3 0.005 41.6
heating, Sub-hour Iron Middle 18 12 2 2 1 – 3 0.01 90
cooling appliance, Dishwasher Middle 20 14 3 3 1.13 – 2 0.015 33.5
integrated-battery Washing machine Middle 18 12 2 2 0.40 – 4 0.01 78.1
devices Washer dryer Middle 18 12 2 2 0.8 – 4 0.01 15.3
Vacuum Middle 18 12 3 3 2 – 4 0.01 93.7
Electric shower Low 18 12 4 4 9 – 12 0.01 67
None Console Middle 24 18 12 12 0.12 – 1 0.005 70.2
Entertainment
Television High 24 18 12 12 0.12 – 1 0.005 99
devices
Computer High 18 12 4 4 0.14 – 1 0.005 70.8
Electric battery Middle 24 18 12 12 0.12 – 6 0.005 22
Electric vehicles, High/
Central heating Middle 18 12 3 3 0.40 – 3 0.01 53
thermal loadings Hour
Electrical panel Middle 18 12 4 4 0.50 – 4 0.01 60

send its own PV generation to the aggregator, which is responsible for will participate in the aggregated DR programme by aggregating its
both, demand reallocation of PV energy provided by prosumers (𝐺𝑃 𝑉 ) preferences for the next day’s energy supply, with the possibility to send
and delivery of energy from the energy service provider (𝐺𝑅𝑊 ). the ‘‘on-site’’ generated renewable energy to the aggregator.
The consumers are classified according to their renewable energy
∑
23
generation (i.e., called prosumers), the possible storage and the re- 𝑡
𝐺𝑅𝑊 = (𝐺𝑃𝑡 𝑉 ,𝐶1 − 𝐷𝐶1
𝑡
)
,𝐶1
quired demand. While certain prosumers (1 ) have an ‘‘on-site’’ renew- 𝑡=0
(2)
able energy generation unit without any storage device, others (2 ) ∑
23
𝑡
have both renewable energy generation and storage facilities. 𝐺𝑅𝑊 ,𝐶2
= (𝐺𝑃𝑡 𝑉 ,𝐶2 + 𝑆𝑃𝑡 𝑉 ,𝐶2 − 𝐷𝐶2
𝑡
)
𝑡=0
Initially, prosumers 1 check their available on-site renewable en-
ergy thanks to the capabilities of the home controller (see Fig. 1). In 𝑆𝑃𝑡 𝑉 ,𝐶2 is the energy supply available by the storage system of pro-
case of enough renewable provision for the entire time slot, consumers sumers 2 that is charged by the surplus 𝑃 𝑉 energy of those prosumers.
can self-supply. If the micro-generated renewable energy is limited When the total energy demand at time 𝑡 exceeds the microgeneration
to meet the whole demand in the entire time horizon, the consumer capability (𝐷𝑇𝑡 > 𝐺𝑃𝑡 𝑉 ), the user will participate in the DR programme

4
C. Cruz et al. Energy Policy 182 (2023) 113745

Table 2
Occupancy patterns for different aggregation scenarios.
Class Occupancy pattern Assumptions
Members with a part-time job in the evenings
Scenario I 6:00–11:00 am
or single pensioners and flexible preferences.
6:00–12 am and Adults with full-time jobs and non pensioner, with energy consumption spread
Scenario II
17:00–23:00 throughout the day and three consumption peaks. Their preferences are not flexible.
6:00–12 am and Working people and multiple pensioners with flexible and fixed
Scenario III
17:00–23:00 demand preferences. Daily energy consumption is distributed in two main periods.

The timeslots follow the occupancy patterns obtained from Dunbabin et al. (2015). The timeslot interval of 6:00–11:00 am (Scenario I) has
been selected to test the reallocation capacity in a context of a high flexibility demand pattern.

and the aggregator will schedule its energy demand for the next 24 h. motivate actors to modify their energy consumption behaviour and/or
The daily energy exchange with the aggregator for each group of share more renewable energy.
consumers is calculated in Eq. (2). |24 |24
𝐺𝑃𝑡 𝑉 ,𝑖 | + 𝑆𝑃𝑡 𝑉 ,𝑖 |
|0 |0
𝑖 = (4)
2.3. Demand reallocation problem formulation |24 |24 |24
𝐺𝑃 𝑉 ,𝑖 | + 𝐺𝑃 𝑉 ,𝑖 | + [𝐺𝑅𝑊 − 𝐷𝑇𝑡 ]|
𝑡 𝑡 𝑡
|0 |0 |0
The aggregator verifies on a daily basis that the total energy de- 3. Results
manded by all appliances of all consumers in the system meets the daily
energy provided by both the service provider and the consumers. The The simulation results are illustrated by conducting experiments
data collector is responsible for a reallocation of the total community on several scenarios to evaluate and analyse the proposed system
demand that is fair to all consumers and at the same time targets the reaction, even in critical cases (i.e. the case of energy outage or higher
available renewable supply for each time slot. peak hours consumption). The reallocated demand mainly affect the
The proposed solution concept to this reallocation problem is given computation time and the reallocation capacity, which depends of
according to Eq. (3). Function 𝐹 handles the optimal reallocation for different demand preferences and conditions (i.e., integration of the
the variable demand according to the time interval for which the device prosumers into the scheduling process). Section 3.1 highlights the
is scheduled. The demand reallocation follows a meta-heuristic method reallocation capability also for the case when the microgeneration is
based on the particle swarm optimisation technique, which provides included at household level. In addition, Section 3.2 presents the results
low computational cost in terms of required time and number of on the reallocation based on the reputation factor. Finally, Section 3.3
iterations needed to find the optimal solution (Kennedy and Eberhart, highlights the regulation and acceptance analysis of aggregated DR in
1995). Therefore, the solutions must satisfy the 𝑡𝑠𝑐ℎ𝑒𝑑,𝑖𝑗 of each device Spain.
in each household and avoid overconsumption in specific time periods.
3.1. Aggregated demand reallocation analysis including microgeneration
𝐷𝑓𝑡 𝑖 + 𝑚𝑖𝑛{ (𝐷𝑣𝑖
𝑡 𝑡
)} ≤ 𝐺𝑅𝑊 𝑡
+ 𝐺𝑅𝑊 ,𝐶1
𝑡
+ 𝐺𝑅𝑊 ,𝐶2
(3)

The 24-hour output vector of scheduled community consumption is In this section, the algorithm behaviour is analysed depending on
given by considering 𝐺𝑅𝑊 and 𝐺𝑃 𝑉 taking into consideration the de- the consumers’ preferences. Those parameters (i.e., 𝑡𝑏𝑒𝑔 , 𝑡𝑒𝑛𝑑 , and the
mand preferences provided by the participants. Function 𝐹 searches the operation time) generate a demand profile with a symmetric demand
optimal time interval for the activation of each device 𝑗 in household 𝑖 trend over 24 h or a larger fluctuation trend in small time slots. Table 2
given its activation time, its preference interval and the available sup- summarises the occupancy patterns and assumptions under three possi-
ply of renewable energy. In particular, taking into account the variables ble scenarios, which are devised to show the effects on the aggregator
, 𝑡𝑠𝑐ℎ𝑒𝑑,𝑖𝑗 , 𝑡𝑏𝑒𝑔,𝑖𝑗 and 𝑡𝑒𝑛𝑑,𝑖𝑗 , the reallocation will determine whether energy management process. The reallocation is performed centrally by
the aggregator, which manages to balance the load uniformly along the
a setting is appropriate by minimising the total overconsumption (in
day. The scenarios are identified as follows:
hours) of the community’s appliances versus the available renewable
Scenario I conforms to a flexible community, where consumers’
supply in a given time interval.
time preferences are relaxed; scenario II follows a rigid community,
where participants have more constrained time flexibility to shift their
2.4. Reputation factor
demand. This scenario is exemplified for a set of consumers with partic-
ular socio-economic or work situations. Finally, Scenario III is focused
We define the reputation factor 𝑅 in which the aggregator will be for a diverse community, where demands display more heterogeneous
able to reliably allocate the available renewable energy. The idea is to dynamism.
identify coalitions in terms of their behaviour as they have patterns in The aggregated DR capabilities are analysed when including pro-
favour of cooperation. The behaviour has been determined on three sce- sumers in the reallocation process. It also specifies how the reallocated
narios, which have been previously defined in Cruz et al. (2021). The demand is influenced under controlled scenarios. The analysis is fo-
reputation factor 𝑅 also helps to distinguish and validate consumers’ cused on the satisfied demand capacity and the needed fossil resources
behavioural patterns. For the proposed framework, the aggregator for three types of situations: (1) consumers send their preferences to the
maintains a reputation value for each individual household based on aggregator under enough 𝐺𝑅𝑊 ; (2) the aggregator owns limited 𝐺𝑅𝑊 to
the amount of renewable energy previously shared. The reputation of supply the whole community; and (3) the microgeneration is included
consumer 𝑖 represents the ratio between the total amount of 𝐺𝑃 𝑉 or in different contexts of demand preferences and 𝐺𝑅𝑊 resources.
𝑆𝑃 𝑉 shared by prosumers, and the available energy demand after the Fig. 2 shows the demand to be reallocated (blue) that is scheduled
scheduling process 𝐺𝑅𝑊 𝑡 - 𝐷𝑇 (Eq. (4)). The reputation factor  helps by a group of consumers in the community, the renewable availability
to validate consumers’ behaviour. It provides a positive value between provided by the energy provider (green) and the reallocated demand
0 and 1 that are dynamic according to each household cooperative achieved by the aggregator (red) under flexible preferences (scenario I).
and energy consumption behaviour. The more renewable energy a The prosumer integration is also evaluated into the scheduling process
household shares, the higher its reputation factor will be. This could (black). Next, Fig. 3 illustrates the reallocated demand achieved under

5
C. Cruz et al. Energy Policy 182 (2023) 113745

Fig. 2. Total reallocated demand in scenario I for enough 𝐺𝑅𝑊 (a); limited 𝐺𝑅𝑊 supply (b), prosumers integration (c). (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)

Fig. 3. Total reallocated demand in scenario II for enough 𝐺𝑅𝑊 (a); limited 𝐺𝑅𝑊 supply (b), prosumers integration (c). (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)

Fig. 4. Total reallocated demand in scenario III for enough 𝐺𝑅𝑊 (a); limited 𝐺𝑅𝑊 supply (b) prosumers integration (c). (For interpretation of the references to colour in this
figure legend, the reader is referred to the web version of this article.)

rigid preferences (scenario II). Finally, Fig. 4 illustrates the reallocation instance, a 3% of the reallocated demand is exceeded for scenario II
achieved by heterogeneous preferences (i.e. users defined by flexible if 10 or more additional prosumers participate in the process. This
and fixed preferences), which is exemplified by the scenario III. It is trend is also shown for the scenario III (1%) (see Fig. 5). The second
worth noting that the amount and type of 𝐺𝑅𝑊 and 𝐺𝑃 𝑉 supply (green) assumption is exemplified in the case of limited 𝐺𝑅𝑊 energy supply
in these scenarios and cases are determined by the service provider and (see Figs. 2b, 3b, 4b). The aggregation DR system tries to maximise the
the capacity of microgeneration is provided by prosumers. The main use of available 𝐺𝑅𝑊 , which is evaluated using the parameter Demand
results are summarised as follow: Reallocation Capacity (DRC). DRC is defined as the ratio of average
The first assumption is stated under enough 𝐺𝑅𝑊 supply (green) load scheduling to the maximum standard deviation per slot obtained
over the entire time slot (see Figs. 2a, 3a, 4a). From those figures, it is after reallocation and it is expressed in percentage format. For instance,
shown that the aggregation is able to supply the required demand of DRC is increased up to 35% in scenario I, but nonetheless, limited
the community (blue) without a need for fossil energy resources. The fossil energy is still needed to supply the entire community (23%).
demand is reallocated according to consumer preferences by reducing The results also show a reduction in peak demand up to 70% in some
peak consumption (red in Fig. 2a). Under enough 𝐺𝑅𝑊 , the 𝑃 𝑉 supply scenarios (see Fig. 2b, scenario I). This situation also arises in scenario
does not have a significant effect on the reallocated demand. For III (heterogeneous preferences), as the community cannot be supplied

6
C. Cruz et al. Energy Policy 182 (2023) 113745

Table 3
DRC and reputation factor 𝑅 obtained for three scenarios with different 𝐺𝑅𝑊 conditions.
Scenario I Scenario II Scenario III
Reputation factor 𝑅 DRC (%) Reputation factor 𝑅 DRC (%) Reputation factor 𝑅 DRC (%)
Limited 𝐺𝑅𝑊 0.68 ± 0.13 59 ± 12 0.50 ± 0.15 23.6 ± 2.2 0.50 ± 0.1 39.4 ± 5
Enough 𝐺𝑅𝑊 0.65 ± 0.15 41.5 ± 9 0.72 ± 0.04 9.3 ± 4.6 0.44 ± 0.12 11.7 ± 0.9

et al., 2021). Pilot experiences focused on the prosumers’ participation

at the household level will facilitate the diffusion of self-production and
the viability of these systems. There are examples to modify this trend,
such as the study proposed by Mir-Artigues and del Río (2021), which
promotes strict self-supply due to the fact that the price of exported
kWh is lower than its generation cost.

3.2. Reputation analysis

Traditionally, a load aggregator maintains cooperative relationships

with participants. They only registers with the aggregator and partici-
pate in the activities. For instance, it is assumed 0.5 kWh as the required
demand to be reallocated by 0.2 kWh available 𝐺𝑅𝑊 . The energy
demand is provided without any priority for the consumers and/or pro-
sumers. In this scenario, consumers who do not participate in electricity
generation could obtain <<green>> energy while prosumers would re-
ceive fossil energy resources. Therefore, the aggregated systems should
recognise consumer behaviour to reallocate energy efficiently, in order
Fig. 5. DRC and fossil resources needed for different scenarios and cases. to encourage the coalition for the installation of renewable generation
and, in turn, to participate optimally in the aggregation process. In
literature, consumption-shift scenarios based on reputation factors are
with only 𝐺𝑅𝑊 and a 19% percentage of fossil resources is required. established to achieve cost savings, where prosumers share their surplus
In this scenario, the two main consumption peaks are reduced due to renewable energy (AlSkaif et al., 2017). To this end, the degree of
the increase of reassigned capabilities up to 50% (see Fig. 4b). The energy sharing can be determined by assigning a reputation factor.
last assumption is exemplified for a 𝑃 𝑉 supply that is included in the Here, experiments have been conducted on prosumers with cer-
aggregated DR system: Figs. 2c, 3c, 4c show the reallocated demand tain reputation value participating in the aggregation turn-process.
achieved when a group of prosumers send their generated 𝑃 𝑉 to the DR Table 3 summarises the 𝑅 factor obtained for three scenarios and 𝐺𝑅𝑊
aggregator. The reallocated demand in this case is properly scheduled conditions, and Fig. 6 illustrates the DRC achieved and the 𝑅 factor
to match the PV renewable supply, specially for the scenario I (see obtained along different days. 𝑅 value is high for prosumers with
Fig. 2c). flexible preferences (scenario I) (𝑅 > 0.5). The result obtained for the
Fig. 5 shows the DRC achieved in the different scenarios. Each scenario I leads to a better management of renewable resources. 𝑅
situation is evaluated based on 𝐺𝑅𝑊 and 𝐺𝑃 𝑉 availability. Under value is reduced for prosumers who require a demand that cannot be
limited 𝐺𝑅𝑊 , DRC presents the highest capacity. However, the ag- easily reasigned (𝑅 = 0.5) regardless of the 𝐺𝑅𝑊 conditions (scenarios
gregator needs extra fossil resources to reallocate all the scheduled II and III). Finally, a one-week analysis is performed to determine the
demand. Enough 𝐺𝑅𝑊 provisions always achieves a positive effect on reputation factor evolution. The scenario II (rigid preferences) offers a
the reallocation process, increasing DRC by 35% or 10% for flexible worse reputation mainly in the early days (see Fig. 6b). However, there
or mixed preferences, respectively. The worst situation is denoted is a slight trend throughout the week concerning 𝑅 improving its value
in scenarios II and III under limited 𝐺𝑅𝑊 supply. The reallocation a few days later (𝛥𝑅 = 0.3 for scenario I,II; and 𝛥𝑅 = 0.5 for scenario
is achieved to mitigate peak consumption taking into account 𝐺𝑅𝑊 III).
and 𝐺𝑃 𝑉 supply. However, the presence of fossil resources appears
to a greater extent for a sufficient community supply (up to 19%).
3.3. Reallocation time analysis in the DR aggregation process
In addition, 𝑃 𝑉 provision improves reallocation capacity in cases of
limited 𝐺𝑅𝑊 supply. For example, the most satisfactory situation is
obtained in scenario I, which increases demand reassignment by 28% The global behaviour is analysed taking into account different cases
to mitigate 𝐺𝑅𝑊 scarcity through prosumers integration. In this case, in the community as follows: type of energy demand (a), trends in
additional fossil energy is hardly needed for certain time slots. demand preferences (b), 𝐺𝑅𝑊 provision (c) and type of behaviour pat-
The increase of prosumers under enough 𝐺𝑅𝑊 does not optimise the tern consumption (d). The Ratio of the Computation Time (RCT) is the
reallocated demand due to the impossibility to achieve the provisions to ratio between the mean required time and the maximum time required
a dis-par time slot. This is a limitation of stand alone 𝑃 𝑉 systems, which for each case, and it is expressed in percentage. It is noticed that the
are considered as non dispatchable renewable energy supply sources differences in the required time for the reallocated demand in cases of
and hence are not able to reallocate the demand generation to other high/low demand are negligible For instance, the algorithm requires
time slots. This can be solved using other complementary technologies, additional resources in a high demand and fluctuations context (RCT
such as energy storage systems. However, Celik et al. (2020) and Ter- = 53,66% - 52.48%). A slight trend is noticed when the aggregator
louw et al. (2019) demonstrate that 𝑃 𝑉 generation negatively affects receives a non-uniform1 energy profile from the service provider 𝐺𝑅𝑊
the aggregator’s profit, as the investment in a storage unit is currently (RCT = 50.34%).
not cost-effective when the lifetime of the battery and its degradation
are being considered. In addition, consumers without on-site energy
1
technologies benefit more than owners of 𝑃 𝑉 storage systems (Zakeri The amount of renewable energy in each of the 24 slots is different.

7
C. Cruz et al. Energy Policy 182 (2023) 113745

Fig. 6. DRC and 𝑅 obtained along the week for three scenarios with enough 𝐺𝑅𝑊 (a); limited 𝐺𝑅𝑊 (b).

3.4. Regulation and acceptance analysis of aggregated DR in Spain frameworks for end-users participation in such programmes based on
environmental, economic or social benefits. The Spanish energy market
Despite the existence of some projects based on smart communities, is aiming for a reduction in consumption peaks and a promotion
explicit regulation has not yet started and demand aggregation is not of flexibility measures at the consumer level to achieve reduction
implemented in the Spanish electricity system. The aggregated schemes in greenhouse gas emission neutrality by 2050 (Spain, 2020). The
are centred for the uninterruptible demand programmes, which are development of energy storage systems and other ICT developments
deployed by large industrial electricity consumers, such as construction for the optimisation of energy consumption and production, which,
industries or factories connected to the high voltage grid. The structure together with the use of other sectors (e.g. electric mobility and electric
of the Spanish electricity sector is regulated and the government is heating) put the consumer at the centre of the energy transition. With
responsible for establishing grid remuneration methods. the introduction of the proposed rules, aggregated DR systems could be
The Spanish regulations and rules approved, Royal Decree-Law a reality in Spain from 2022, year to carry out qualification tests and
(RDL) 900/2015 introduced a slow transition towards a distributed market participation. We are undoubtedly at a crucial moment for the
generation model considering small-scale systems. Furthermore, RDL development of new energy management systems.
15/2018 removed the so-called solar tax to give certainty to energy pro- An initial analysis of aggregated DR acceptance is carried out
sumers. The regulation defines different types of self-consumption: col- in Cruz (2022), which was distributed telematically via web. Conceived
lective and individual. In addition, consumers, individually or through as a preliminary study, it aims to measure the impact and acceptance
an aggregator, can provide all energy needs at any time of the day for both, efficiency platforms and microgeneration capabilities on
to the service provider.2 This regulatory framework aims to create a aggregated DR systems for residential consumers. The survey is detailed
more efficient and dynamic market as a whole, in which the consumer in Surv (2022), and the main results are summarised below:
becomes an active element in the market and establishes measures
1. The majority of the participants (97%) show interest in the
to promote energy efficiency. Measures to deploy the participation of
knowledge and/or use of energy efficiency tools. However, most
small consumers were considered in the electricity RDL 24/2013 and
of them are not familiar with concepts such as smart energy
RDL 216/2014. In turn, the BOE-A-2021-8362 legislation resolved on
communities, aggregation or DR (17%), and IoT-based user ap-
May 6th 2021 already approved the operating rules for the daily and
plications (30%). In terms of renewable resources, a low per-
intra-daily electricity markets. More specifically, the 11.a rule empha-
centage of interviewees are not familiar with the generation and
sises the review of procedures’ needs for the storage and the aggregator
management of resources at home (23%). This result may be due
incorporation. These actors will not be linked to the electricity sale.
to a lack of familiarity with ICT - IoT capabilities applied in the
On other hand, the integrated national energy and climate plan
residential sector. Specifically, a limited number or participants
2021–2030 presented by Institute for energy diversification and sav-
would be interested in the use of renewable energy resources
ing (Idae, 2020) highlights the need for renewable resources integration
(37%) and to a lesser extent, users would be willing to use
into the electricity system. At the same time, RDL 23/2020 is focused
microgeneration resources (20%).
on measures for the promotion of renewable energies. It anticipates
2. In general, consumers understand peak generation (72%) and
accessibility regulation and connection issues including action mecha-
demand and/or price could reduce consumption in the future.
nisms based on the ‘‘pay-as-bid’’ system. In order to achieve the climate
However, only a 60% of participants are interested in appliances
objectives assumed by Spain, a whole regulatory development is still
automatisation and a lower percentage (37%) would be willing
pending.3
to plan and/or estimate their energy demand for the next day.
The technical developments and pilot implementation can be seen
This low acceptance is mainly due to a lack of awareness of
as a proof of concept and a starting point for establishing regulations
the benefits for a shared energy resource system through the
for cooperative DR programmes. This will require involvement of other
use of ICTs. One of the causes may be due to concerns about
staekholders in the energy sector and careful development of legal
the security of communications and the real benefit of using
aggregated DR in their daily lives (40%).
2
Markets publication and competition regulation was performed by the 3. Consumers would be fully involved in preferences modification
national commission (18423/2019). in order to organise the consume (54%). On the contrary, only
3
The regulation should define the relationship model between the service 11% of participants would demand consumption time prefer-
provider and the aggregator. ences with no scope for demand reallocation. In general, users

8
C. Cruz et al. Energy Policy 182 (2023) 113745

are in favour of applications to schedule their preferences (95%) 4.2. Policy implications
and would have no problem installing a controller in their
home to control their appliances (90%), activating/deactivating Firstly, the residential sector is very sensitive to energy price signals.
appliances according to the reprogrammed demand. Therefore, the economic implications could focus on proposals for
replacing price signal structures and, where appropriate, introduce
In literature, the studied surveys suggest security and privacy as
additional incentives or policy mechanisms to support energy carriers
essential factors to implement a smart DR infrastructure (Vardakas
of greater efficiency. One of the main weaknesses in policy implications
et al., 2015). It is also recommended to explore money-saving op-
of DR systems is the unstable effectiveness of policy instruments and
portunities as reasons for changing electricity consumption patterns.
the commitment. It is due to the low quality or under-implementation
However, there is evidence that households will revert to their old
of instruments, the poor enforcement, or the ineffective instrument
consumption habits if they cannot see the impact of their behavioural
design.
changes on their daily lives (Annala et al., 2014). Therefore, it is
Secondly, tackling climate change requires a change in patterns and
important to provide consumers with meaningful information about
methods of consumption. Policies, as well as new legal and regula-
demand reallocation processes in order to encourage sustainable en-
tory frameworks should be put in place to create incentives for new
ergy consumption (Ellabban and Abu-Rub, 2016). Social influence,
return energy demand structures. Relevant concepts, such as storage,
social and personal awareness to changes needed in the society, habit
aggregation and renewable communities, were recently introduced by
formation, individual self, will help stimulate sustainable consumer
the Spanish RDL 23/2020, which adopted measures in the field of
behaviour changes (White et al., 2019).
energy and other areas of economic recovery from COVID-19. However,
4. Conclusions and policy implications Spain has not considered energy efficiency and flexibility policies (RDL
6/2022) and even it has not developed directives concepts that were
4.1. Conclusions defined in RDL 23/2020.
Thirdly, a very close coordination in the planning and actions
The development of efficient mechanisms for energy management in execution between the public institutions and end demand consumers
smart communities is still an open challenge. Residential communities would be essential in order to be able companies and participants make
tend to play an important role to the success of DR programmes. decisions in a rational and efficient manner. This action requires the
The consumption measurement and energy use, as well as instilling introduction of specific rules (i.e. options on how consumers can partic-
energy-efficient behaviours in consumers, are as necessary as promot- ipate in the energy market). To this regard, RDL 244/2019 put in place
ing advances on energy use. We envision the aggregated DR framework clearer definitions of prosumers, simplified compensation schemes, and
to evolve towards achieving a balanced method between green energy streamlined technical and administrative requirements. It is expected
consumption and provision. Proposed as an optimal demand realloca- to reduce previous barriers to entry for self-consumption, as energy
tion scheme for a smart residential community, the role of aggregation service providers can offer the optimal technical solutions.
for enabling distributed energy resources is to provide electricity ser- Finally, if the electrical grid is decarbonised, the residential housing
vices at scale. The research is focused on the analysis of the consumers’ sector can meet the 28% emission reduction target for 2025 under the
behaviour by analysing the performance in terms of demand reallo- Paris Agreement (Pan et al., 2017). The large-scale use of renewable
cation capacity and computation time not only for consumers, but energies requires the mobilisation of significant investments in genera-
also for proactive consumers that voluntarily participate in the ag- tion or pilot DR infrastructures, among others. A CO2 tax or a floor price
gregated process. To this regard, the inclusion of microgeneration would provide a clear price signal for the reduction of CO2 emissions
technologies based on renewable energy plays an important role when by new funds to contribute to new technologies (e.g. aggregated DR,
limited renewable energy is supplied by the service provider, as each home storage implementation and renewables production).
consumer can assign individually its demand, or by the aggregator. Spain has shown important leadership on clean energy transitions
However, the unlimited increase of prosumers participation does not regulations. When all Spanish plans and strategies are implemented, a
necessarily optimise the reallocation resources (the DRC value obtained completely different energy sector will emerge, where fossil fuels are no
remains constant). Furthermore, the reputation factor helps us to anal- longer dominant. In conclusion, there are good reasons for governments
yse three types of consumer behaviour, specially for households that to strive to give subsidies to smart solution platforms and hardware
are equipped with 𝑃 𝑉 systems and flexible preferences. needed to realise cooperative DR systems. It is also important to take
The results of a preliminary analysis to uptake aggregated DR a comprehensive approach for market measures and system designs,
systems are also presented. It is concluded that consumers would be weighing the different challenges and striving to find a balance that
willing to participate in a collaborative DR system, but they would speed up the transition to a more sustainable and smart residential
be concerned about its security and functionality operation. This re- energy sector.
search work can be enhanced by analysing the main services that an
aggregation process can offer to end-users in order to stimulate a more CRediT authorship contribution statement
responsible energy use. One of the main limitations is the difficulty
of extracting consumer flexibility information from the data set used Carlos Cruz: Conceptualization, Methodology, Investigation, Writ-
for validation of the proposed system. Furthermore, pilots deployment ing – original draft, Writing – review & editing. Tarek Alskaif: In-
in living labs are needed to validate the results of our analysis. We vestigation, Writing – review & editing. Esther Palomar: Investiga-
have defined an aggregation model that can be adapted to the Spanish tion, Funding acquisition, Writing – review & editing. Ignacio Bravo:
electricity market. However, there is a need for determining the roles, Investigation, Funding acquisition, Writing – review & editing.
rights and responsibilities of independent aggregators and end-users.
This can be accelerated by conducting pilot projects in a controlled Declaration of competing interest
environment. The development of a coordination system between ag-
gregators, market operators and customers should be considered in The authors declare that they have no known competing finan-
order to avoid distortions and undesired consequences on the current cial interests or personal relationships that could have appeared to
system. A future study of how the prosumers could be compensated for influence the work reported in this paper.
their participation will be performed. An even greater challenge to be
performed is the real-time power generation application and demand Data availability
estimation methods that may also be employed to adjust upcoming
operation and properly schedule consumer demands. No data was used for the research described in the article.

9
C. Cruz et al. Energy Policy 182 (2023) 113745

Fig. 7. Average energy demand of 29 matching households profiles with no children and multiple pensioners members in scenario I (a); single and non pensioner members’
matching 35 households in scenario II (b); and average energy demand of 259 matching heterogeneous households profiles in scenario III (c). The mean of behavioural profiles
was designed by Architectural Research as part of DECC and DEFRA’s analysis of the Household Electricity Study (Dunbabin et al., 2015).

Acknowledgements Chukhnina, A., Melnikov, G., Pecherkin, A., Sokolov, A., Yanovich, Y., 2021. Reputation
system for local flexibility market. In: 2021 3rd Conference on Blockchain Research
Applications for Innovative Networks and Services. BRAINS, pp. 4–5. http://dx.doi.
This paper is a result of the ENEFF-PILOT project, funded by Co- org/10.1109/BRAINS52497.2021.9569815.
munidad de Madrid, Spain (Spain) under the Talent Attraction Grant Conchado, A., Linares, P., Lago, O., Santamaría, A., 2016. An estimation of the
2017, with reference 2017-T1/TIC- 5184. economic and environmental benefits of a demand-response electricity pro-
gram for Spain. Sustain. Prod. Consump. 8, 108–119. http://dx.doi.org/10.
1016/j.spc.2016.09.004, URL: https://www.sciencedirect.com/science/article/pii/
Appendix S2352550916300239.
Crespo-Vazquez, J.L., AlSkaif, T., González-Rueda, Á.M., Gibescu, M., 2020. A
We include as an Appendix a more detailed simulation assumptions community-based energy market design using decentralized decision-making under
uncertainty. IEEE Trans. Smart Grid 12 (2), 1782–1793.
of the scheduling process. Fig. 7 shows the mean occupancy patterns
Cruz, C., 2022. A Cooperative Approach to Aggregated Demand Response Schedul-
and type of energy demand for three possible scenarios. ing: Sustainable Communities Optimizing Renewable Energy Use (Ph.D. thesis).
University of Alcalá, unpublished.
Cruz, C., Palomar, E., Bravo, I., Aleixandre, M., 2021. Behavioural patterns in
References
aggregated demand response developments for communities targeting renewables.
Sustainable Cities Soc. 72, 103001. http://dx.doi.org/10.1016/j.scs.2021.103001,
AlSkaif, T., Lampropoulos, I., van den Broek, M., van Sark, W., 2018. Gamification- URL: https://www.sciencedirect.com/science/article/pii/S2210670721002857.
based framework for engagement of residential customers in energy applications. Cruz, C., Palomar, E., Bravo, I., Gardel, A., 2019. Towards sustainable energy-efficient
Energy Res. Soc. Sci. 44, 187–195. http://dx.doi.org/10.1016/j.erss.2018.04.043, communities based on a scheduling algorithm. Sensors 19 (18), http://dx.doi.org/
URL: https://www.sciencedirect.com/science/article/pii/S2214629618304420. 10.3390/s19183973, URL: https://www.mdpi.com/1424-8220/19/18/3973.
AlSkaif, T., Luna, A.C., Zapata, M.G., Guerrero, J.M., Bellalta, B., 2017. Reputation- Dunbabin, P., Palmer, J., Terry, N., 2015. Electricity use by domestic appliances in
based joint scheduling of households appliances and storage in a microgrid English households. In: ECEEE-2015 European Council for an Energy Efficient
with a shared battery. Energy Build. 138, 228–239. http://dx.doi.org/10. Economy. pp. 1439–1450.
1016/j.enbuild.2016.12.050, URL: https://www.sciencedirect.com/science/article/ Eds, 2022. Garantía en el accesso a una energía asequible, segura, sostenible y
pii/S0378778816319314. moderna para todas las personas. Archit. [online]. https://www.agenda2030.gob.
AlSkaif, T., Zapata, M.G., Bellalta, B., 2015. A reputation-based centralized energy es/objetivos/objetivo7.htm. (Access 01 May 2022).
allocation mechanism for microgrids. In: 2015 IEEE International Conference on Ellabban, O., Abu-Rub, H., 2016. Smart grid customers’ acceptance and engagement:
Smart Grid Communications. SmartGridComm, pp. 416–421. http://dx.doi.org/10. An overview. Renew. Sustain. Energy Rev. 65, 1285–1298. http://dx.doi.org/10.
1109/SmartGridComm.2015.7436336. 1016/j.rser.2016.06.021, URL: https://www.sciencedirect.com/science/article/pii/
Ambrosio-Albala, P., Upham, P., Bale, C., Taylor, P., 2020. Exploring acceptance of S1364032116302441.
decentralised energy storage at household and neighbourhood scales: A UK survey. EPA, 2021. Agencia protección ambiental Estados Unidos. protec. [online].
Energy Policy 138, 111194. http://dx.doi.org/10.1016/j.enpol.2019.111194, URL: https://www.usa.gov/espanol/agencias-federales/agencia-de-proteccion-
https://www.sciencedirect.com/science/article/pii/S0301421519307797. ambiental-de-estados-unidos. (Access 29 June 2021).
Annala, S., Viljainen, S., Tuunanen, J., Honkapuro, S., 2014. Does knowledge contribute Faria, P., Barreto, R., Vale, Z., 2019. Demand response in energy communities
to the acceptance of demand response? J. Sustain. Dev. Energy Water Environ. Syst. considering the share of photovoltaic generation from public buildings. In: 2019
2 (1), 51–60. International Conference on Smart Energy Systems and Technologies. SEST, pp.
Bauwens, T., Schraven, D., Drewing, E., Radtke, J., Holstenkamp, L., Gotchev, B., 1–6. http://dx.doi.org/10.1109/SEST.2019.8849082.
Yildiz, Ö., 2022. Conceptualizing community in energy systems: A systematic Fiestras-Janeiro, M.G., García-Jurado, I., Meca, A., Mosquera, M.A., 2011. Cooperative
review of 183 definitions. Renew. Sustain. Energy Rev. 156, 111999. http://dx. game theory and inventory management. European J. Oper. Res. 210 (3), 459–466.
doi.org/10.1016/j.rser.2021.111999, URL: https://www.sciencedirect.com/science/ Gallego-Castillo, C., Heleno, M., Victoria, M., 2021. Self-consumption for energy
article/pii/S1364032121012624. communities in Spain: A regional analysis under the new legal framework. Energy
Bibri, S.E., Krogstie, J., 2017. Smart sustainable cities of the future: An exten- Policy 150, 112144. http://dx.doi.org/10.1016/j.enpol.2021.112144, URL: https:
sive interdisciplinary literature review. Sustainable Cities Soc. 31, 183–212. //www.sciencedirect.com/science/article/pii/S0301421521000136.
http://dx.doi.org/10.1016/j.scs.2017.02.016, URL: http://www.sciencedirect.com/ Guo, X., Guo, X., 2015. China’s photovoltaic power development under policy
science/article/pii/S2210670716304073. incentives: A system dynamics analysis. Energy 93, 589–598.
Brugger, H., Eichhammer, W., Mikova, N., Dönitz, E., 2021. Energy Efficiency Guo, Z., Zhou, K., Zhang, C., Lu, X., Chen, W., Yang, S., 2018. Residential
Vision 2050: How will new societal trends influence future energy demand electricity consumption behavior: Influencing factors, related theories and inter-
in the European countries? Energy Policy 152, 112216. http://dx.doi.org/10. vention strategies. Renew. Sustain. Energy Rev. 81, 399–412. http://dx.doi.org/10.
1016/j.enpol.2021.112216, URL: https://www.sciencedirect.com/science/article/ 1016/j.rser.2017.07.046, URL: http://www.sciencedirect.com/science/article/pii/
pii/S0301421521000859. S1364032117311164.
Camarinha-Matos, L.M., 2016. Collaborative smart grids – A survey on trends. Renew. Haidar, N., Attia, M., Senouci, S.-M., Aglzim, E.-H., Kribeche, A., Asus, Z.B.,
Sustain. Energy Rev. 65, 283–294. http://dx.doi.org/10.1016/j.rser.2016.06.093, 2018. New consumer-dependent energy management system to reduce cost
URL: https://www.sciencedirect.com/science/article/pii/S1364032116303264. and carbon impact in smart buildings. Sustainable Cities Soc. 39, 740–750.
Celik, B., Suryanarayanan, S., Roche, R., Hansen, T.M., 2020. Quantifying the impact http://dx.doi.org/10.1016/j.scs.2017.11.033, URL: http://www.sciencedirect.com/
of solar photovoltaic and energy storage assets on the performance of a residential science/article/pii/S2210670717307047.
energy aggregator. IEEE Trans. Sustain. Energy 11 (1), 405–414. http://dx.doi.org/ Hallen, J.G., 2021. Healthy Buildings. Housh. [online]. https://www.hsph.harvard.edu/
10.1109/TSTE.2019.2892603. joseph-allen/healthy-buildings/. (Access 29 June 2021).

10
C. Cruz et al. Energy Policy 182 (2023) 113745

Idae, 2020. Ministerio para la transición ecológica y el reto demográfico. PNIEC. Powells, G., Bulkeley, H., Bell, S., Judson, E., 2014. Peak electricity demand and
[online]. https://www.idae.es. (Access 30 June 2021). the flexibility of everyday life. Geoforum 55, 43–52. http://dx.doi.org/10.1016/
IEA, 2022. Example of daily load profile for solar PV production relative to electricity j.geoforum.2014.04.014, URL: https://www.sciencedirect.com/science/article/pii/
demand in 2050, IEA, Paris. IEA. [online]. https://www.iea.org/. (Access 26 July S0016718514000931.
2021). Sauter, R., Watson, J., 2007. Strategies for the deployment of micro-generation:
Ine, 2021. Spanish statistical office. INE. [online]. https://www.ine.es. (Access 30 Implications for social acceptance. Energy Policy 35 (5), 2770–2779. http:
March 2022). //dx.doi.org/10.1016/j.enpol.2006.12.006, URL: https://www.sciencedirect.com/
Inês, C., Guilherme, P.L., Esther, M.-G., Swantje, G., Stephen, H., Lars, H., 2020. Regula- science/article/pii/S0301421506004903.
tory challenges and opportunities for collective renewable energy prosumers in the Shareef, H., Ahmed, M.S., Mohamed, A., Al Hassan, E., 2018. Review on home
EU. Energy Policy 138, 111212. http://dx.doi.org/10.1016/j.enpol.2019.111212, energy management system considering demand responses, smart technologies, and
URL: https://www.sciencedirect.com/science/article/pii/S0301421519307943. intelligent controllers. IEEE Access 6, 24498–24509. http://dx.doi.org/10.1109/
Iria, J., Soares, F., Matos, M., 2018. Optimal supply and demand bidding strat- ACCESS.2018.2831917.
egy for an aggregator of small prosumers. Appl. Energy 213, 658–669. http:// Silva, B.N., Khan, M., Han, K., 2018. Towards sustainable smart cities: A review of
dx.doi.org/10.1016/j.apenergy.2017.09.002, URL: https://www.sciencedirect.com/ trends, architectures, components, and open challenges in smart cities. Sustainable
science/article/pii/S0306261917312758. Cities Soc. 38, 697–713. http://dx.doi.org/10.1016/j.scs.2018.01.053, URL: http:
Islam, T., Islam, R., Pitafi, A.H., Xiaobei, L., Rehmani, M., Irfan, M., Mubarak, M.S., //www.sciencedirect.com/science/article/pii/S2210670717311125.
2021. The impact of corporate social responsibility on customer loyalty: The Silviu, C., 2022. Legislative train schedule Fit for 55 package under the European Green
mediating role of corporate reputation, customer satisfaction, and trust. Sustain. Deal. FIT55. [online]. https://www.europarl.europa.eu/committees/en/the-energy-
Prod. Consump. 25, 123–135. http://dx.doi.org/10.1016/j.spc.2020.07.019, URL: aspects-of-the-fit-for-55-pac/product-details/20211026CHE09623. (Access 06 May
https://www.sciencedirect.com/science/article/pii/S2352550919304944. 2022).
Jovanovic, R., Bousselham, A., Bayram, I.S., 2016. Residential demand response Singh, S., Yassine, A., 2019. Mining energy consumption behavior patterns for house-
scheduling with consideration of consumer preferences. Appl. Sci. 6 (1), http://dx. holds in smart grid. IEEE Trans. Emerg. Top. Comput. 7 (3), 404–419. http:
doi.org/10.3390/app6010016, URL: https://www.mdpi.com/2076-3417/6/1/16. //dx.doi.org/10.1109/TETC.2017.2692098.
Kashanizadeh, B., Shourkaei, H.M., Fotuhi-Firuzabad, M., 2022. A bi-stage multi- Souza, A.T., Milbradt, R.G., Canha, L.N., Silva Santana, T.A., 2018. Development of
objectives optimization of smart multi energy system with active consumers. a demand response system integrated to photovoltaic microgeneration and energy
Sustain. Prod. Consump. http://dx.doi.org/10.1016/j.spc.2022.03.028, URL: https: storage using IoT and artificial intelligence. In: 2018 15th International Conference
//www.sciencedirect.com/science/article/pii/S2352550922000811. on the European Energy Market. EEM, pp. 1–5. http://dx.doi.org/10.1109/EEM.
Kennedy, J., Eberhart, R., 1995. Particle swarm optimization. In: Proceedings of 2018.8469971.
ICNN’95 - International Conference on Neural Networks, Vol. 4. pp. 1942–1948 Spain, D., 2020. Integrated National Energy and Climate Plan 2021–2030. DGEG,
vol.4. http://dx.doi.org/10.1109/ICNN.1995.488968. Madrid, Spain.
Koecklin, M.T., Longoria, G., Fitiwi, D.Z., DeCarolis, J.F., Curtis, J., 2021. Public Surv, 2022. Comunidades inteligentes, eficientes y sostenibles. SURV. [online]; (Access
acceptance of renewable electricity generation and transmission network de- 06 May 2022). http://www3.uah.es/eneffpilot.
velopments: Insights from Ireland. Energy Policy 151, 112185. http://dx.doi. Terlouw, T., AlSkaif, T., Bauer, C., van Sark, W., 2019. Multi-objective optimiza-
org/10.1016/j.enpol.2021.112185, URL: https://www.sciencedirect.com/science/ tion of energy arbitrage in community energy storage systems using different
article/pii/S0301421521000549. battery technologies. Appl. Energy 239, 356–372. http://dx.doi.org/10.1016/
Kramers, A., Höjer, M., Lövehagen, N., Wangel, J., 2014. Smart sustainable cities j.apenergy.2019.01.227, URL: https://www.sciencedirect.com/science/article/pii/
– Exploring ICT solutions for reduced energy use in cities. Environ. Model. S0306261919302478.
Softw. 56, 52–62. http://dx.doi.org/10.1016/j.envsoft.2013.12.019, URL: http: Torriti, J., 2017. Understanding the timing of energy demand through time use data:
//www.sciencedirect.com/science/article/pii/S136481521400019X. Thematic issue Time of the day dependence of social practices. Energy Res. Soc. Sci. 25, 37–
on Modelling and evaluating the sustainability of smart solutions. 47. http://dx.doi.org/10.1016/j.erss.2016.12.004, URL: https://www.sciencedirect.
Liu, L., Li, B., Qi, B., Qin, Q., Sun, Y., Chen, S., Zhu, C., 2020. Reliable interoperation of com/science/article/pii/S2214629616302985.
demand response entities considering reputation based on blockchain. IEEJ Trans. Vardakas, J.S., Zorba, N., Verikoukis, C.V., 2015. A survey on demand response pro-
Electr. Electron. Eng. 15 (1), 108–120. grams in smart grids: Pricing methods and optimization algorithms. IEEE Commun.
Lo Piano, S., Smith, S., 2022. Energy demand and its temporal flexibility: Approaches, Surv. Tutor. 17 (1), 152–178. http://dx.doi.org/10.1109/COMST.2014.2341586.
criticalities and ways forward. Renew. Sustain. Energy Rev. 160, 112249. http://dx. Wang, Z., Li, H., Deng, N., Cheng, K., Lu, B., Zhang, B., Wang, B., 2020. How to
doi.org/10.1016/j.rser.2022.112249, URL: https://www.sciencedirect.com/science/ effectively implement an incentive-based residential electricity demand response
article/pii/S1364032122001708. policy? Experience from large-scale trials and matching questionnaires. Energy
López Prol, J., Steininger, K.W., 2020. Photovoltaic self-consumption is now profitable Policy 141, 111450. http://dx.doi.org/10.1016/j.enpol.2020.111450, URL: https:
in Spain: Effects of the new regulation on prosumers’ internal rate of return. //www.sciencedirect.com/science/article/pii/S0301421520302032.
Energy Policy 146, 111793. http://dx.doi.org/10.1016/j.enpol.2020.111793, URL: White, K., Habib, R., Hardisty, D.J., 2019. How to SHIFT consumer behaviors to be
https://www.sciencedirect.com/science/article/pii/S0301421520305140. more sustainable: A literature review and guiding framework. J. Mark. 83 (3),
Mir-Artigues, P., del Río, P., 2021. Prosumers’ behavior under a regulation that en- 22–49.
courages strict self-sufficiency. The case of spanish photovoltaic micro-generation. Yamaguchi, Y., fei Chen, C., Shimoda, Y., Yagita, Y., Iwafune, Y., Ishii, H., Hayashi, Y.,
Energies 14 (4), http://dx.doi.org/10.3390/en14041114, URL: https://www.mdpi. 2020. An integrated approach of estimating demand response flexibility of domestic
com/1996-1073/14/4/1114. laundry appliances based on household heterogeneity and activities. Energy Policy
Pan, X., den Elzen, M., Höhne, N., Teng, F., Wang, L., 2017. Exploring fair and 142, 111467. http://dx.doi.org/10.1016/j.enpol.2020.111467, URL: https://www.
ambitious mitigation contributions under the Paris Agreement goals. Environ. sciencedirect.com/science/article/pii/S0301421520302184.
Sci. Policy 74, 49–56. http://dx.doi.org/10.1016/j.envsci.2017.04.020, URL: https: Zakeri, B., Gissey, G.C., Dodds, P.E., Subkhankulova, D., 2021. Centralized vs. dis-
//www.sciencedirect.com/science/article/pii/S1462901117302502. tributed energy storage – Benefits for residential users. Energy 236, 121443.
Parrish, B., Heptonstall, P., Gross, R., Sovacool, B.K., 2020. A systematic re- http://dx.doi.org/10.1016/j.energy.2021.121443, URL: https://www.sciencedirect.
view of motivations, enablers and barriers for consumer engagement with com/science/article/pii/S0360544221016911.
residential demand response. Energy Policy 138, 111221. http://dx.doi.org/10.
1016/j.enpol.2019.111221, URL: https://www.sciencedirect.com/science/article/
pii/S0301421519308031.

11