Renewable Energy 221 (2024) 119732

Contents lists available at ScienceDirect

Renewable Energy
journal homepage: www.elsevier.com/locate/renene

Optimal integration of hybrid pumped storage hydropower toward

energy transition
Helena M. Ramos , Jeremy E. Sintong , Alban Kuriqi *
CERIS, Instituto Superior Técnico, University of Lisbon, 1649-004, Lisbon, Portugal

A R T I C L E I N F O A B S T R A C T

Keywords: This study explores the advantages of combining variable renewable energy sources like solar and wind with a
Clean energy pumped storage hydroelectric (PSH) system for grid integration. The hybrid modeling systems considered in this
Energy crisis study consist of four distinct schemes and seasons to ensure their adaptability to real-world conditions. Each
Energy policy
scheme is defined by specific rules and algorithms, including those without PSH, with PSH, with PSH adjust­
Energy transition
Renewable energy
ments, and with an optimized schedule for the PSH storage system. The last scheme of daily simulation with an
optimization solver is used to optimize the integration of PSH with a hybrid power system that uses solar and
wind energy as primary renewable sources by minimizing the daily operating costs. The optimal value is derived
from the minimized operating costs and incoming and outgoing energy accumulation. The dispatch system in­
cludes load demand satisfaction considering the intermittent nature of solar and wind sources and demand
fluctuations presented with a hybrid system of PV solar, wind turbines, PSH, and other power generators as
backup. Limitations of the study include considering various factors influencing the integration of a PSH unit,
such as different ecological and geographical conditions, distinct atmospheric data, and varying load and de­
mand patterns, which may result in different outcomes in different situations. The results show that using the
developed model to optimally schedule the integration of PSH with renewables and the hybrid system in each
season provides significant cost savings and CO₂ emissions reductions while meeting the same load patterns.
Namely, it can be inferred from the tested schemes that PSH plays a crucial role in facilitating the transition to
sustainable energy sources. This is evident through project financing results, which indicate a favorable positive
Net Present Value (NPV) and a relatively short payback period of 5 years. Furthermore, Scheme 4 demonstrates
the potential for a substantial 84 % reduction in CO₂ emissions during the summer season, and daily costs can be
significantly lowered by over 90 % across all seasons. Overall, two simplified net present value and payback
period estimation models have shown feasible project financing for a simulation period of 25 years with an
interest rate of 10 % for the combined PSH and hybrid system.

1. Introduction
Energy storage systems allow the storage of surplus energy during
periods of high generation and low demand and deliver energy to the
power grid during periods of high demand when energy production is
insufficient to meet demand [1]. The world does not currently have
sufficient energy storage—and the storage that does exist is almost
exclusively pumped hydroelectric plants operating in tandem with hy­
droelectric plants on rivers. Therefore, we need more energy storage to
support solar and wind power [2]. Grid-connected batteries are suitable
for short-term storage of minutes to hours. However, pumped storage is
excellent for overnight and longer-term storage [1]. PSH is the most
efficient technology for energy storage systems and requirements [3].
Solar photovoltaic (PV) and wind turbines (WT) should be integrated
with energy storage systems to enable a clean energy transition and use
energy from renewable sources more efficiently [4,5]. The global
installed PSH capacity was 158 GW at the end of 2019 [6]. PSH is a
widely used and proven energy storage technology, accounting for 93 %
of the world’s energy storage capacity. There are 130 pumped storage
power plants in 42 countries worldwide and more than 40 existing
plants in the United States [3]. The main reasons for the massive
development of PSH are the growing demand for fluctuating renewable
energy (VRE) storage, the increasing demand for peak power, and the
modernization of plants to improve efficiency [7].

* Corresponding author.
E-mail addresses: helena.ramos@tecnico.ulisboa.pt (H.M. Ramos), jeremysilalahi.je@gmail.com (J.E. Sintong), alban.kuriqi@tecnico.ulisboa.pt (A. Kuriqi).

https://doi.org/10.1016/j.renene.2023.119732
Received 6 September 2023; Received in revised form 3 November 2023; Accepted 23 November 2023
Available online 30 November 2023
0960-1481/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
H.M. Ramos et al.
Despite some limitations and low market mechanisms, PSH com­
pensations in the grid include primary frequency response and voltage
control [8]. However, the future of PSH will depend on this advanced
integration in the grid, which creates new revenue streams. Energy
storage is essential for energy arbitrage and ancillary services and re­
quires flexible control [9]. The energy system charges storage when
energy is abundant and cheap and returns or discharges it when energy
supply is scarce and expensive compared to demand. Conversely, the
auxiliary services usually refer to the operating reserves of active power.
The importance of energy storage is a reality. It is also accelerating as
more and more countries have committed to using renewable energy as a
major component of their stimulus programs to achieve net zero emissions
[10]. In 2020, the Intergovernmental Panel on Climate Change found that
energy production contributes to more than two-thirds of global green­
house gas emissions [11]. Therefore, transitioning to green and clean en­
ergy is fundamental to achieving net zero by 2050. Globally installed solar
and wind power capacity has increased from about 1 GW of solar PV and
17 GW of wind power in 2000 to about 650 GW each, followed by solid
government subsidies [12]. However, according to the [13], more than
250 TWh of intermittent renewable electricity was curtailed in 2020,
almost equal to Spain’s annual electricity demand. Had this wasted clean
energy production been stored for later use, about 180 million tons (MT) of
CO2 emissions could have been avoided [14]. Minimizing curtailment and
wasted energy also improves renewables’ utilization and capacity factor,
which is expected to yield a higher return on investment.
A key point for electricity storage that supports its economic perspec­
tive is the ability to derive multiple value streams by enabling a range of
services with the proper storage system application [15]. This will improve
project returns and allow for the stacking of revenue streams. Changes are
needed in some places, including changes in project schedule structure and
regulations and new markets for ancillary grid services [2]. In addition,
behind-the-meter applications are needed to access utility markets
through aggregators and maximize the storage contribution. Alternatively,
potential cost savings from energy storage systems combined with the shift
toward optimizing energy curtailment and economic dispatch. PSH leads
the world in installed electricity storage capacity, accounting for 96 % of
the 176 GW (GW) installed worldwide. Other types of electricity storage
technologies already in significant use worldwide include thermal storage
at 3.3 GW (1.9 %), batteries at 1.9 GW (1.1 %), and other mechanical
storage at 1.6 GW (0.9 %) [16].
Comparison of energy storage technologies consisting of PSH,
lithium-ion battery (LFP), lead-acid battery, vanadium redox flow bat­
tery (VRF), compressed air energy storage (CAES), and bidirectional
hydrogen energy storage system, with a system size of 100 MW for 4 h
(except hydrogen for 10 h) [17,18]. The technologies, except hydrogen,
have more than ten years of references in commercial operation [19].
Pumped storage has a better lifetime, excellent response time from
standstill to whole generation in about 2–5 s, outstanding technical
capabilities (i.e., technological readiness, the inertia for grid resilience),
2000€/kW average electricity CapEx, ~500€/kWh for average energy
CapEx. A comparison of energy storage technologies for 100 MW, 4 h
runtime in 2020 & 2030 shows an effective CapEx of 2700€/kW, which
is lower than lithium-ion battery storage (4570), lead-acid batteries
(5070), CAES compressed air (3300), and hydrogen (9000) [18]. Pro­
jections for 2030 show values for effective investment costs of
2700€/kW, which are lower than LI -ion battery storage (3200€/kW),
lead-acid batteries (4000€/kW), CAES compressed air (3300€/kW), and
hydrogen (4620€/kW) [20,21]. A similar scenario is used to compare
energy storage technologies for 1.000/100 MW, 10 h duration in 2020
and 2030, considering the effective investment cost based on a PSH
lifetime of 80 years and a discount rate of 6 %, leading to a PSH of
2900€/kW, which is lower than Li-ion Battery storage (~8100€/kW),
Lead Acid Battery (~9000€/kW), CAES compressed air (~3100€/kW)
and hydrogen (~4600€/kW) [20,21].
It can be concluded that PSH solutions provide a better overall lifetime
than batteries alone [1]. The estimation of energy storage technologies is
2
Renewable Energy 221 (2024) 119732
based on the value of the initial end-of-life investment, including the cost
of replacement at the end of each lifetime. Not only because batteries have
limited capacity to store large amounts of renewable energy
cost-effectively but also because they wear out due to limited cycling and
need to be replaced regularly, typically every 5–15 years [1]. In contrast,
PSH systems have between 50 and 100 years of life.
In a power system, it is always essential that production is at least
equal to or higher than demand. According to Ref. [16], about 86 % of
electricity will come from renewable sources by 2050. The 60 % would
come from solar and wind, with an installed capacity of over 6000 GW
(wind) and 8500 GW (solar). This considerable increase in intermittent
renewables will upset the global power grid planning and operations
context. Electric supply and demand must be balanced through reliable
sources and low-carbon alternatives [22].
Nowadays, hydropower is the primary producer of renewable elec­
tricity. In total, the installed capacity of hydropower has reached 1330
GW (i.e., over 17 % of the total global electricity generation), producing
a record 4370 TWh of clean electricity in 2020 [23]. To keep the tem­
perature rise below 2 ◦ C, an additional 850 GW of hydropower capacity
is needed by 2050, including pumped storage [16]. To limit temperature
rise to no more than 1.5 ◦ C, even more hydropower capacity must be
added, which is considered technically feasible, cost-effective, and so­
cially acceptable. In this scenario, renewables generate 88 % of total
electricity; wind and solar have a share of 68 %, and the contribution of
hydropower is about 10 %. With a total renewable capacity of 26,600
GW, hydropower would have at least 2660 GW installed in 2050 [13].
According to the IEA’s net-zero scenario [13], total hydropower ca­
pacity would double by 2050, adding at least 1300 GW. This model
shows the importance of combining the holistic hydropower energy
system with pumped storage in a broader mix of renewable sources to
bridge the gap and complement each other. The new PSH development,
especially in a system with increased wind and solar capacity, would
improve the reliability and flexibility of the system while reducing the
need to build a new fossil-based power plant [7].
Flexible and more reliable production is an essential prerequisite for
integrating intermittent energies. In the case of PSH, when the other
energy sources produce too much power, the turbine pumps the water in
reverse mode to the higher reservoir to function like a giant natural
battery and is available again when needed. When demand is high,
pumped storage will be critical to match production to consumption
constantly - the most adaptable production system. Hydropower will
continue to play a crucial role in the energy transition [6]. Pumped
storage will be a large-scale energy storage solution, providing reactive
power for frequency regulation, synchronous or virtual inertia, and
black start capability [18]. Solar and wind energy are intermittent re­
sources that fluctuate daily and from hour to hour. They may not match
the required energy consumption (i.e., load or demand) [3]. According
to the simulation results, maximizing Pumped Storage Hydropower
(PSH) use in ancillary service markets (specifically, spinning and/or
non-spinning services) is more profitable than focusing solely on seeking
arbitrage opportunities [9]. PSH can participate in energy spinning
reserve in pumping mode and non-spinning reserve offline mode. PSH
can also be optimized to minimize the cost over the entire horizon [2].
The PSH is the answer to the best possible prices for energy and
ancillary services. In a liberalized energy market, the market for ancil­
lary services such as products (other than energy) is supported by the
PSH to maintain the capacity and transmission of energy from resources
to loads and ensure the power system’s reliable operation and power
quality [9]. It also plays a crucial role in frequency regulation and
instantaneous grid balancing.
Demand side management (DSM) includes load shedding that
shuts–down loads in the industrial and residential sectors, peak shaving,
or load shifting [24]. PSH is very well suited as a complementary partner
to other renewables. When electricity from renewable sources (i.e., wind
and solar) exceeds demand, pumped storage provides large-scale energy
storage that intercepts the excess supply because it can be stored for later
H.M. Ramos et al.
use whenever needed. In a liberalized electricity market, where the
share price fluctuates with high demand, pumped storage can help sta­
bilize electricity prices by providing electricity at an affordable price due
to its reliability and ramp-up capability [25]. It is the best available
technology for storing energy in large quantities to support the inte­
gration of renewable energy.
The primary purpose of this study is to demonstrate the valuable
benefits of integrating the PSH system into the power grid in combina­
tion with hybrid systems with renewable energy sources. Therefore, this
study is based on the following questions: it is necessary to systemati­
cally demonstrate why the presence of PSH as an integrated storage
system in the existing grid is inevitable. To answer the central question
of how a given amount of energy can be stored at minimum cost with
optimal scheduling, modeling of the integration of PSH systems with
other renewable energy sources is developed. In addition, how PSH
actively contributes to reducing the global carbon footprint of the en­
ergy sector by supporting the integration of significant amounts of wind
and solar energy will be explored. Another urgency of PSH deployment
in this project involves a viable scheme of PSH power system to reduce
the high price of the power grid from a diesel generator by absorbing and
storing excess renewable energy and meeting peak demand during high
power consumption. It also analyzes how this model can prove the
effectiveness of PSH deployment in the hybrid system.
This study is divided into the main concepts of the PSH to the main
results of integrating the PSH into the power system with the modeling of a
hybrid system. Section 2 presents the PSH data and assumptions, the
developed model, and system boundary conditions and details the oper­
ating principles of the PSH for which a closed-loop model was chosen.
Section 3 presents the technical analysis of the optimization of PSH
scheduling with intermittent renewables and describes the other operating
components accordingly. It includes a complete description of 4 schemes
to compare the costs of hybrid systems, each in different seasons. Section 4
shows the valuable benefits of PSH in terms of CO₂ reduction and project
financing. It also includes estimating CO₂ emissions from the hybrid sys­
tem model and the simplified project financing of PSH integration with the
hybrid system model. Section 5 presents a comprehensive display of the
simulation results. Section 6 delves into the significance of the study’s
findings in the context of related literature. Finally, Section 7 summarizes
the main findings resulting from this study.
2. Materials and methods
2.1. Data and PSH operation assumptions
We gather parametric data from various reputable sources to repli­
cate real-world modeling procedures. The representative hourly load or
demand data comes from REN, the only transmission authority in
Portugal, and refers to 2019. The parametric information supporting the
weather conditions for renewable energy comes from the Azores, a small
island in the Atlantic Ocean that belongs to Portugal. The corresponding
weather data comes from a web tool developed by Imperial College
London and ETH Zurich called Renewables. ninja (https://www.renew
ables.ninja). Electricity generation from solar and wind energy follows
basic calculations involving solar and wind turbine technologies
selected in a specific datasheet from the representative manufacturers of
the technology providers.
The hybrid modeling systems are divided into four schemes and
seasons to maintain the robustness of other influenceable parameters in
real life. Each scheme has rules and algorithms that can be described as
without PSH, with PSH, with adjusted PSH, and with optimized sched­
uling of the PSH storage system. The different schemes allow each to
operate in pumping or turbine mode, depending on the demand and
resources in a given hour. The results are discussed annually and daily,
considering some constraints.
This study focuses exclusively on the grid integration of a hybrid
system with PSH as a complementary storage system to renewables.
3
Renewable Energy 221 (2024) 119732
Some associated challenges related to external factors in operating a
hybrid system (i.e., environmental and hydrology considerations) are
neglected because the PSH tested in this study is a closed-loop system
not connected with natural aquatic ecosystems such as rivers or lakes.
The analyses were performed based on the following assumptions:
• First, the closed PSH loop is used for the schematic analysis of the
primary system operation.
• The energy system with an integrated pump and turbine calculator
flexibly adjusts even at a fixed speed, automatically adjusting the
grid frequency in either the pump or turbine flow direction. This
power system allows for increased revenue by selling more power at
peak demand at the highest possible price. In other words, mini­
mizing costs maximizes the benefits of PSH integration with the
hybrid plant.
• In addition, pumped storage is assumed to have no losses due to
evaporation or leaks in the pipelines and reservoirs.
• The PSH schematic model is integrated to rapidly ramp up genera­
tion during peak periods as the available hydro fixed resources ca­
pacity further expands variable renewable energy (VRE). Thus, the
flow rate follows the power generation and varies dynamically per
hour, depending on the adjustment scheme.
• Emission reduction is based on a developed algorithm that simulta­
neously reduces the total applied cost.
2.2. PSH working principles and closed-loop model
Pumped storage power plants use water in two different reservoirs
(upper and lower basins), separated by a particular head, to store po­
tential energy. Excess electricity from the grid or intermittent renewable
energy sources (e.g., solar and wind power) can be used during periods
of low demand to pump water from the lower reservoir into the upper
reservoir, turning it into a large battery storage system [26]. Two water
reservoirs are located at different elevations and can generate elec­
tricity. The water flows through a turbine and generates electricity by
pumping it into the upper reservoir.
When demand increases, water from the upper reservoir flows
through turbines to generate electricity. This method allows flexibility
in meeting dynamic demand due to seasonal conditions or weather de­
pendencies when the sun is not shining or wind speed is insufficient to
turn on the wind turbines [6]. The electricity generated is reliable
because it is based on the well-known principle of hydropower, which
produces large amounts of energy over a long period, thus allowing
generation that can be switched off.
Based on natural inflows from water sources, PSHs can be divided into
closed-loop and open-loop configurations. The closed-loop configuration
includes two reservoirs isolated from any free-flowing water source; the
reservoirs can be artificial or natural [27]. In contrast, an open loop system
has a permanent hydrologic connection to a natural water body.
The open-loop system always has a portion of electricity that can be
generated without pumping, as in a conventional hydroelectric system.
In the closed-loop category, the upper reservoir is located outside the
river. On the other hand, the open-loop has an upper reservoir that is
usually upstream, and both reservoirs may have natural inflows. The
closed loop is often preferred because it less impacts the environment,
such as fish passage or sediment migration. The modeling includes the
closed-loop concept in this research [27].
3. Optimization of PSH scheduling with intermittent renewables
3.1. Case study and operation components
Solar and wind energy are intermittent resources that naturally
fluctuate daily and hourly. They may not match ultimate residential or
industrial energy use. A typical annual load is obtained from available
load data from Ref. [28], the electricity transmission company in
H.M. Ramos et al. Renewable Energy 221 (2024) 119732

Fig. 1. Pumping and turbine mode operation of PSH.

Portugal that has a concession contract with the Portuguese state. The wind turbine manufacturer’s power curve. On the other side, rated wind
corresponding source refers to 2019 data. The main purpose is to obtain speed, Urated is the nominal wind speed on the optimal operation of wind
the load pattern; therefore, a reduction multiplication factor 50 was turbine generation [29]. The maximum allowable power output is
applied to simulate the hourly demand on a 1-year basis. The dynamic limited to the rated wind speed. In the pumping mode, the grid feeds
data consists of solar radiation (kW/m2), wind speed (m/s), ambient electricity to the motor (Grid-In). In contrast, the generator injects
temperature (◦ C), and air density (kg/m3), all retrieved from Renew­ electricity into the grid (Grid-Out) during turbine mode, as seen in Fig. 1.
ables. ninja (https://www.renewables.ninja). A specific point in Portu­ The associated equation of turbine mode can be described as follows,
guese territory in the Azores with a latitude of 37.8085 and a longitude Eq. (6):
of − 25.4731 was chosen.
Pturb. =ηturb. x ρ(water) x g x Q x Hnet (6)
The highest and lowest recorded wind speeds were 24.85 and 0.14
m/s, respectively, while solar irradiance was 1.009 and 0 kW/m2, While during pumping mode, the equation is applied as follows, Eq.
respectively. The most expensive recorded hourly energy price was (7):
180.3 €/MWh, while the cheapest was 12.15 €/MWh. Moreover, based
on initial calculations, the average hourly load/demand was higher than
ρ(water) x g x Q x Hnet
Ppump. = (7)
the average hourly PV and wind generation, corresponding to 114.95 ηpump.
and 92.32 €/MWh, respectively. Therefore, the PSH system was devel­ Maximum stored energy in the upper reservoir (UR) corresponds to
oped to ensure that hourly operations could be managed to balance the net head of the hydropower system, written as in Eq. (8):
supply and demand while minimizing operating costs. The calculations
for photovoltaic power plants are described in equations (1) and (2). EUR− max =ρ x Vmax x g x Hnet (8)

GT (t) ( ( )) The power supply must always be greater or at least the same as the
PPV (t) = PRated PV x x 1 + kT x TC (t)− Tref (1) power demand (or load). The hourly demand equation can be described
Gref
as follows, Eq. (9):
TC (t)=Tamb (t)+0.03 GT (t) (2)
PDemand or Load (t) ≤ PPV(t) +PWT(t) +Pturbining(t) − Ppumping(t) +PGrid− In(t) − PGrid− Out(t)

kT is the temperature coefficient of solar PV modules and TC is cell’s (9)

temperature. The dynamic data has given different temperatures on an Operational restrictions of solar photovoltaics (PV), wind turbines
hourly basis, and so the TC changes based on ambient temperature and (WT), grid-in, grid-out, PSH units, and storage capacity are determined
solar irradiance. Tref is accustomed to reference temperature during test as follows, Eq. (10):
conditions (assumed to be 25 ◦ C) while Tamb is the ambient temperature
(in Kelvin). The calculations for wind power plant are described in 0 ≤ PDemand or Load (t) (1 ≤ t ≤ N)
equations (3)–(5).
[ ] 0 ≤ PPV(t) (1≤ t ≤ N)
1
PWT (t)=ηWT x x CP x ρa (t)x π x R2 x U 3 (t) if Ucut− In ≤ U(t) ≤ Urated (3)
2 0 ≤ PWT(t) (1≤ t ≤ N)
[ ]
1 0 ≤ PGrid− In (t) (1≤ t ≤ N)
PWT (t)=ηWT x x CP x ρa (t)x π x R2 x Urated
3
(t) if Urated ≤ U(t) ≤ UCut− Out
2
(4) 0 ≥ PGrid− Out (t) (1≤ t ≤ N)

PWT (t)= 0 if U(t) ≤ UCut− In or U(t) ≥ UCut− Out (5) 0 ≤ Pturbining(t) (1≤ t ≤ N)

In that case, ηWT , the wind turbine’s efficiency usually ranges from 0.2 0 ≤ Ppumping(t) (1≤ t ≤ N)
to 0.4, depending on wind data and wind turbine types. CP is the power
coefficient, and its value is always less than 0.59, complying with Betz VRmin ≤ VR(t) ≤ VRmax (10)
Limit’s theorem. The area of a wind turbine is calculated from the
The reservoir simplified model is carried out to put constraints on
diameter of the rotor blade, R, and its value is always constant, while on
PSH operation, represented in Fig. 2.
the other side, air density varies per hour, ρa (t). UCut− In is wind speed
The maximum storage capacity is less than the maximum volume, i.e.,
lower bound for the turbine to start rotating and hence producing
the full supply level (FSL). At the same time, the initial state of pumping
power, while UCut− Out is the wind speed upper bound at which the
(SoPInit) is set at 50 % of FSL. Following the abovementioned restrictions,
generator is decoupled from the shaft, and the turbine stops producing
the minimum storage capacity is 30 % of the maximum capacity.
power. This applicable computation is a standard practice based on the

4
H.M. Ramos et al. Renewable Energy 221 (2024) 119732

the objective functions each year. The different schemes are compared
daily in simulations.

3.2.1. Scheme 1 (without PSH)

Fig. 2. Reservoir operation simplified model.
3.2. System integration and optimization
The integrated output of the PV system, wind turbine, grid feed-in,
grid feed-out, and PSH system is shown in Fig. 3. Grid injection is
equivalent to diesel generation, which is always used as a last resort to
balance the deficit between energy supply and demand.
Moreover, its contribution to the grid should be minimized because
the higher the diesel production, the higher the CO₂ emissions generated
by fossil fuel combustion. During excess energy, the system will supply
power to the engine to pump water from the lower reservoir to the upper
reservoir. When there is insufficient energy to meet demand, the addi­
tional energy is provided by directing the water into the lower reservoir
of the PSH and driving the turbine generator.
The reservoir catchment area is constant so that the parametric
volume rate can be monitored by reservoir operation. The stored energy
can be calculated from the law of potential energy if the contained
volume, gravity, and given net head are known. Efficiency is lower in
pumping mode than in turbine mode because the shape of the impeller is
reversed from turbine to pumping mode. Fig. 3 shows that the hybrid
power system has been integrated into the PSH system. Four schemes
were developed to make technical comparisons to find the best way to
schedule the PSH into the grid and determine the appropriate values of
In scheme 1, no PSH is used. Instead, renewable generation de­
termines whether electricity is injected into the system from the grid, i.
e., grid-in, or whether excess produced electricity is exported from the
system to the grid, i.e., grid-out. The following formulas determine the
equations, equations (11)–(14):
( )
Grid Determination=PDemand or Load (t) − PPV(t) +PWT(t) (11)
( ( ))
PGrid− In(t) = If PDemand or Load (t) − PPV(t) +PWT(t) > 0 (12)
( ( ))
PGrid− Out(t) = If PDemand or Load (t) − PPV(t) +PWT(t) < 0 (13)
( )
CostApplied (t) = PGrid− In(t) − PGrid− Out(t) x Price(t) (14)
3.2.2. Scheme 2 (with PSH)
In this scheme, the pumped storage mode is enabled and tied
exclusively to the positive value of the State of Pumping (SoP). SoP
determines the ratio between the remaining and the fully charged
pumped capacity. The constraint also includes the maximum storage
capacity for SoP operation. The flowchart representing the schematic
diagram of Scheme 2 can be seen in Fig. 4.
3.2.3. Scheme 3 (with adjusted PSH)
The SoP can be flexibly created in pump or turbine mode in this
scheme. Its positive and negative values are noted as non-constraining
variables. The next iteration of the SoP is adjusted with the previous
time step of SOP, associated with the maximum pumping rate at a given
capacity. In Scheme 3, the SoP is flexibly adapted to decide whether the
hourly operation should be in turbine or pumping mode based on the
maximum pumping rate and the previous state of the SoP. In Scheme 2, the
SoP is only for a value greater than 0 and depends on the maximum storage
capacity at a constant rate. The flowchart of Scheme 3 can be seen in Fig. 5.

Fig. 3. Energy system of pumped storage hydropower integration with intermittent renewables and electrical power grid.

5
H.M. Ramos et al. Renewable Energy 221 (2024) 119732

Fig. 4. Flowchart of operation scheme 2.

Fig. 5. Flowchart of operation scheme 3.

3.2.4. Scheme 4 (with optimized PSH) Subject to (Constraints):

The main objective of Scheme 4 is to minimize the costs associated
PSupply(t) ≥ PDemand or Load(t) (1 ≤ t ≤ N)
with integrating pumped storage power plants into hybrid solutions by
varying the hourly output of pumps, turbines, grid injection, and grid
0 ≤ SOP ≤ SOPmax
(t) (1 ≤ t ≤ N)
withdrawal.

[{( ) } {( )} ]
min PGrid− In (t) +PGrid− Out (t) x Celec.price(t) + Pturbining(t) x Cturbining(t) +Ppumping(t) x Cpumping(t) (15)

6
H.M. Ramos et al. Renewable Energy 221 (2024) 119732

Table 1
Diesel parametric information.
PGrid− Out(t) ≤ Pmax.Grid− Out (t) (1 ≤ t ≤ N)

Diesel Parametric Information Value Dimension By changing parameters:

Low Heating Value (LHV) 44 MJ/kg
Density (ρ(diesel) ) 0.85 kg/L PPumping (X1); PTurbining (X2); PGrid− In (X3); PGrid− Out (X4) (16)
CO₂ Emissions Rate (ERCO2 ) 2.62 kg CO₂/L(diesel)
The simulation was performed using a developed solver based on
1 kWh 3.6 MJ

Table 2
Project finance parametric information for solar PV, wind turbine, turbine, and pumped storage systems.
Project Finance Parametric Information Value Dimension

Solar PV Specific Investment Costa 863 €/kW

Specific Operation & Maintenance Costb 8.63 €/kW/year
Salvage Valuee 32880.3 €
Wind Turbine Specific Investment Costa 1200 €/kW
Specific Operation & Maintenance Costc 15 €/kW/year
Salvage Valuee 5040 €
Hydropower - Turbining Mode (HP) Specific Investment Costa 1870 €/kW
Specific Operation & Maintenance Costd 24.21 €/kW/year
Salvage Valuee 36091 €
Pumped Storage Hydropower – Pumping Mode (PSH) Specific Investment Costf 950 €/kW
Specific Operation & Maintenance Costg 9.5 €/kW/year
Utilization Cost Tariffh 1 €/MWh
Salvage Valuee 18335 €
i
Electricity Feed-in-Tariff P1 (1st – 10th Year) 87.64211 €/MWh
Electricity Feed-in-Tariff P2 (11th – End) j 61.34948 €/MWh
Project Lifetime 25 Years
Interest Rate 10 %

Notes.
a
IRENA [16] with adjusted assumptions and self-weighted justifications.
b
1 % from the Specific Investment Cost of Solar PV.
c
1.25 % from the Specific Investment Cost of Wind Turbine.
d
1.3 % from the Specific Investment Cost of Hydropower.
e
10 % from each installed generation capacity.
f
[23]with adjusted assumptions and self-weighted justifications.
g
1 % from the Specific Investment Cost of PSH.
h
[23] with adjusted assumptions and self-weighted justifications.
i
Yearly Average from [28].
j
70 % from Electricity F.I.T in P1

GRG Non-Linear Optimization. The result is presented graphically to

obtain the optimal hourly scheduling of PSH operation. It also includes
the distribution of the volumetric profile of pumps and turbines, the
profile of the generated energy of pumps and turbines, the total cost
flow, and the total daily power profile, each for four different seasons
(winter, spring, summer, and autumn).

4. PSH valuable benefits of CO₂ reduction and project financing

4.1. Estimation of CO₂ emissions from the hybrid system model

CO₂ is a harmful emission for humans and the environment; there­

Fig. 6. Preliminary yearly load vs. renewables supplied generation.
Pturbining (t) ≤ Pmax.turbining (t) (1 ≤ t ≤ N)
fore, its presence should be quantified based on the grid power gener­
ated [30]. It depends on the type of fuel burned and the emission rate of
the diesel engine. The parametric information for diesel operation can
be found in Table 1.
Since the power fed into the system from the grid varies from time to
time, the calculation of CO₂ emissions is linear to the amount of power
fed into the grid, as described in Eq. (17):
ERCO2 x PGrid− In (t) x 3.6
XCO2 (kg) = (17)
ρ(diesel) x LHV(diesel)
The higher the accumulated grid-in power, the higher the CO₂
emitted to the atmosphere.

Ppumping (t) ≤ Pmax.pumping (t) (1 ≤ t ≤ N)

7
H.M. Ramos et al. Renewable Energy 221 (2024) 119732

Fig. 7. Yearly power (upper figure) and cost distribution profile (lower figure) in scheme 1.

4.2. Simplified project financing of PSH integration with hybrid system
model
A project funding scheme is needed to consider whether the associ­
ated hybrid integration with PSH is financially viable. Regarding project
financing, the output data from the hybrid system model is included in
the simulation of 25 years of assumed mechanical parts of the project life
cycle with an interest rate of 10 %. The other input parameters for
project financing are shown in Table 2.
5. Results
As preliminary data processing, the distribution profile of electricity
generation from PV and wind turbine technologies is plotted annually
against hourly electricity demand. The fluctuation of intermittent
renewables makes it clear that electricity demand cannot be met unless
an energy storage solution is integrated into the system. The corre­
sponding distribution profile is shown in Fig. 6.
The weather behavior variables were adjusted to see the variations in
the four seasons (i.e., spring, summer, fall, and winter), as shown in
Figure A1. It can be seen that the cumulative electricity generation is
higher in summer due to the high intensity of solar radiation [31]. In
winter, on the other hand, demand cannot be met by electricity gener­
ation from wind and solar alone during most periods. There are times
when the electricity supply is much higher than the demand. However,
the load pattern fluctuates in a relatively more predictable profile [32].
For example, the daily load is high during peak hours. This can be
explained by the fact that more electrical energy is needed to power
industry, services, air conditioning, heating, and electronic equipment.
This typical daily load profile co-occurs. Solar and wind energy cannot

Table 3
Summary of daily demand samples and energy production from renewables used in the systems.
Description Winter Spring Summer Autumn

Total Daily Demand – Sample (MWh) 2.27 2.88 2.75 2.82

Total Daily Energy Produced by PV Generation – Sample (MWh) 0.92 1.76 2.99 2.04
Total Daily Energy Produced by Wind Generation – Sample (MWh) 0.65 0.28 0.0065 0.28

8
H.M. Ramos et al.
Fig. 8. Daily power and cost distribution profile during summer and winter for scheme 1.
always meet the hourly demand, even when the sun peaks in summer.
This energy limitation should be considered, so energy storage sys­
tems must be included later in the other systems [3]. In times of excess
energy production, it can be stored when the energy supplied from
renewable sources is not sufficient. The annual energy distribution
profile and applied costs are shown in Fig. 7.
Fig. 7 describes the daily grid-in and distribution cost profile without
the PSH application. Grid-in lines show additional electricity needed
from the grid (i.e., auxiliary diesel generation) to meet the demand that
cannot be met by certain renewables alone [33]. On the other hand,
off-grid lines show excess energy produced by renewables when load/­
demand is met and then injected into the grid. The daily cost profile in
Fig. 7 (bottom figure) shows a positive value when demand is more
dominant than renewably generated power; therefore, more costs are
imposed on the hybrid system. The lower the applied cost, the more
9
Renewable Energy 221 (2024) 119732
profitable the hybrid system is, so a negative value means a profit (i.e., a
surplus of electricity). The preliminary daily data are taken from each
representative date of each season. The summary of the daily samples of
renewable energy demand and generation is shown in Table 3.
In general, the samples of daily demand vary from season to season.
According to the selected samples, electricity consumption is lowest in
winter (2.27 MWh); in spring, electricity consumption is highest among
the other seasons, at about 2.88 MWh per day. The highest daily power
generation from wind occurs in winter, with about 0.65 MWh. In com­
parison, energy generated from photovoltaics peaks at nearly 3 MWh in
summer. Fig. 8 shows the daily electricity and cost distribution profile in
summer and winter based on the preliminary scheme 1. In this scheme 1,
no energy storage system is used to meet the demand. Figs. 8 and 9 are
the daily representative scheme 1, simplifying and comparing the model
with the other schemes. Scheme 4 is the final scheme expected to be the
H.M. Ramos et al.
Fig. 9. Daily power and cost distribution profile during Spring and Autumn in scheme 1.
optimal solution for integrating PSH into the grid. The daily sample is
taken randomly from each representative season and is dedicated
exclusively to daily conditions.
Fig. 8 shows a surplus of electricity during normal working hours due
to excess electricity from renewable sources, mainly solar energy. For
this reason, the daily cost distribution for the corresponding hours is also
negative. However, from evening to early morning, the renewable
power sources are less than the predicted loads [34]. Therefore, addi­
tional power generation outside the hybrid system is required. The more
electricity from the grid is injected into the hybrid system, the more the
total cost is burned while producing more carbon emissions due to
higher diesel consumption [11]. The decision to buy or sell electricity in
Scheme 1 depends solely on the difference between supply and demand.
Fig. 9 describes the daily electricity and cost distribution profile in
the spring and fall based on preliminary scheme 1. The overall profile is
10
Renewable Energy 221 (2024) 119732
almost the same as in Fig. 8, with daily demand not being met after
normal working hours until early morning. Consequently, more addi­
tional power needs to be injected into the hybrid system.
Fig. 10 shows the annual volume and energy distribution profile as
the main distinguishing feature from what is not valid for the daily
distribution profile of Scheme 1. In Scheme 2, volumetric profiles are
presented during pumping and turbine operation due to the inclusion of
the PSH in the hybrid system. However, in some cases, the annual cost
accumulation may vary between Schemes 1 and 2, depending on load
demand and weather profiles.
According to Fig. 10, the pumping state (SoP) is set to a minimum
value of 0, which describes the conditions independent of the previous
state at each iterated time step. This means that the pumping mode is
used only when the pumping energy does not exceed the maximum
storage capacity at a given hour. SoP measures the energy available in a
H.M. Ramos et al.
Fig. 10. Yearly volumetric and energy distribution profile of pumping and
turbine mode in Scheme 2.
pumped storage system at a given hourly condition, varying from hour
to hour. The annual power and cost distribution profile of scheme two is
shown in Fig. 11. Using the defined algorithms of scheme two, it can be
seen that the cost profile of scheme two is biased towards higher prof­
itability due to lower injection into the grid and higher withdrawal from
the grid in the simulation. This has already been stated for PSH as the
chosen energy storage system for daily analysis. Scheme 1 has no stor­
age, while in Scheme 2, PSH is selected as the energy storage system.
Comparing the annual distribution of pumping and turbine energy
between schemes 2 and 3, more reliable operation is achieved in scheme
3, allowing the PSH system to operate with more freedom, as shown in
Figs. 12 and 13. Negative SoP values indicate that the condition of a
given hour depends on the maximum pumping rate and the previous SoP
value. This scheme does not apply to the previous scheme 2, which is the
main difference. Fig. 12 refers to the distribution profile of scheme 3 and
illustrates the above explanation. The comparison between Figs. 12 and
10 shows the difference.
The annual power and cost allocation profile has also been adjusted
in Fig. 13 and follows the daily dynamic hourly demand. For example,
when SoP is needed, Scheme 2 is subject only to the maximum storage
capacity. At the same time, in Scheme 3, the storage decision depends on
11
Renewable Energy 221 (2024) 119732
Fig. 11. Yearly power and cost distribution profile in Scheme 2.
the maximum storage capacity and the state of the previous hour, which
is linked to the forecast for the next hour and the state of the previous
hour. The better accumulation of the cost allocation profile is also
achieved in scheme 3, based on Fig. 13, which will also focus on daily
analysis.
The annual cost profile of Scheme 3 is the same as Scheme 2. How­
ever, the applied cumulative costs are more profitable in scheme 3,
which is fully summarized in the conclusion. The incremental cost of the
distribution profile has improved by adjusting the PSH mode. Schemes
2, 3, and 4 require a slightly different daily cost distribution to be
extracted than Scheme 1. Fig. 14 shows the daily costs, each for the
corresponding season. The lowest cumulative costs are achieved when
the PSH solution is integrated into the system. This allows flexibility in
adjusting the network when pumping from a given load.
The annual cost profile of Scheme 3 is the same as Scheme 2. How­
ever, the applied cumulative costs are more profitable in scheme 3,
which is fully summarized in the conclusion. The incremental cost of the
distribution profile has improved by adjusting the PSH mode. A slightly
different cost distribution from Scheme 1 Schemes 2, 3, and 4 requires a
somewhat different daily cost distribution, extracted using Schemes 2, 3,
and 4 and Scheme 1. Daily cumulative costs, each for the corresponding
season, are presented in Figure A3. The lowest incremental costs are
H.M. Ramos et al.
Fig. 12. Yearly volumetric and energy distribution profile of pumping and
turbine mode in Scheme 3.
obtained when the PSH solution is integrated into the system. This al­
lows a flexible adjustment of the network when pumping from a given
load. As shown in Figures A3 and A4, plans 2 and 3 have similar patterns
in scheduling the use of the PSH system. In the early morning, the de­
mand is met with a hydroelectric turbine, so the volumetric turbine
output decreases. The contribution from wind power is minimal at this
time. Pumping occurs during daylight hours due to the high contribution
of renewable energy from the photovoltaic system.
Figs. 15–17 show the energy distribution profiles for all resources.
Fig. 15 describes the energy distribution of the pumping and turbine
modes within the different seasons in Scheme 2. In Scheme 2, the State
of Pumping (SoP) is strictly limited to the pumping mode once it has
reached maximum storage capacity. Therefore, there is no lower bound
less than 0 during the day. The State of Pumping (SoP) plays a larger role
in the early morning so that wind energy can be used later during peak
demand. At the same time, photovoltaics always has a larger share
during the day. This makes intuitive sense, as each operating strategy
can be adjusted in many ways with different thresholding methods,
whether to use some parts of the wind, solar, or hydro or to use small
parts of it partially while raising the other thresholds to an affordable
level.
12
Renewable Energy 221 (2024) 119732
Fig. 13. Yearly power and cost distribution profile in Scheme 3.
The daily distribution profile in each season of Scheme 3 is shown in
Fig. 16. The SoP can be adjusted in each previous hour, depending on the
state of the SoP in the last hour, allowing greater flexibility with other
renewable sources. The influence of the seasonal aspect does not seem to
be too big, as the pattern looks almost similar. The most significant
common difference is the amount of energy in each hour. For example,
in the summer, SoP puts out much more power than in the winter due to
the large amount of energy produced by solar PV. However, focusing on
meeting demand in Schemes 2 and 3 does not result in the optimal cost
that can be achieved. This is because the price for each hour varies from
time to time, and of course, the variation of the demand leads to
different results.
The minimum cost of PSH scheduling can be derived from each
schedule 4 in each season, as shown in Fig. 17. This is because the main
objective is to minimize the total cost of an integrated system. The
electricity price is much lower in the early morning than during daylight
hours. Therefore, the optimized method for Scheme 4 has a much larger
SoP recorded to be as optimal as possible in the early morning hours.
Even if the demand cannot always be met by photovoltaics, wind
turbines, and hydropower alone, the grid-in (e.g., diesel production) can
be used accordingly [35]. Considering all parametric information with
PSH schemes, the most optimal way to use the applied costs is in Scheme
4. As can be seen from scheme 4, most of the electricity is sold during the
H.M. Ramos et al. Renewable Energy 221 (2024) 119732

Fig. 14. Daily cost distributions within different schemes and seasons.

13
H.M. Ramos et al. Renewable Energy 221 (2024) 119732

Fig. 15. Results of daily distribution profile within different seasons in Scheme 2.

Fig. 16. Results of daily distribution profile within different seasons in Scheme 3.

14
H.M. Ramos et al.
Fig. 17. Results of daily distribution profile within different seasons in Scheme 4.
daytime due to the higher price (Fig. 19). At the same time, other means
of electricity distribution can be operated in hybrid mode, respecting all
constraints. The illustration generated corresponds to the conditional
state of the SoP maximized in the early morning, considering the dy­
namic electricity price.
The distribution pattern of optimal grid integration with PSH can
also vary, depending on the load profile and renewable sources. Figs. 18
and 19 summarize the distribution profile of electricity flow and pum­
ped turbine energy in each scheme across different seasons. The trend
line shows a cost reduction from Scheme 1 to Scheme 4, indicating
optimized scheduling of PSH operation to find the most optimal path to
achieve minimum cost. The daily schematic simulation scale depends on
hourly demand and other associated resources.
In some cases, the applied cost of a scheme is somewhat smaller and
different from the bulk of the distribution, which is a small fraction
(Fig. 19). In this study case, this analysis has shown the optimal way to
integrate PSH into the hybrid renewable energy system, thereby
achieving a gain or minimizing the cost and producing fewer carbon
emissions.
The daily cost in Scheme 1 is the highest among the other schemes,
with the conspicuous absence of a PSH unit. The total pumping and daily
distributed energy distribution in summer is higher than in any different
season (Fig. 19). This can be explained by the fact that more solar ra­
diation and electricity are available during this period.
The ratio of total pumping power to total turbine energy is the
highest of all schemes for scheme 4 in each season. This tendency is also
true for Scheme 3, which is higher than Scheme 2 (Fig. 19). This refers to
the percentage of power stored and later withdrawn, or the electrical
energy generated divided by the electrical energy used to pump the
water. The higher the ratio, the lower the energy loss in storage, given
the demand and arbitrage parameters to minimize the cost function. It
can be concluded that diesel (grid-in) contributions are undervalued in
all systems and seasons with an integrated power system that uses
pumped storage.
15
Renewable Energy 221 (2024) 119732
The less grid-in is accumulated, the less CO₂ emissions are released
into the atmosphere because less diesel fuel is produced (Fig. 20). For
example, without the PSH integration of Scheme 1, CO₂ emissions in the
spring are estimated to be 0.3675 tons/day. At the same time, in Scheme
4, costs are minimized, and the contribution of Grid-In is at the lowest
level of 0.0977 tons.
The accumulation of CO₂ generated in summer is relatively lowest
compared to other seasons because more electricity is generated from
photovoltaic systems during daylight hours and summertime. The
annual report in Fig. 20 shows that the use of PSH units in the hybrid
system has significantly contributed. The annual CO₂ reduction from
scheme 1 to scheme 3 is 28.4 %. On the other hand, the daily analysis
within the different seasons also shows a significant reduction of diesel
generation by grid injection by operating the PSH units in an optimal
schedule.
Nevertheless, the project financing scheme has shown a positive NPV
value, meaning the projected investment returns exceed the expected
costs. With a project duration of 25 years and an interest rate of 10 %, it
can be assumed that the investment will pay for itself by the 5th year.
The assumptions considered the operating schedule of all the units
connected to the hybrid system, combining the hours in which the
photovoltaic and wind turbines are in operation with the yield according
to the schedule of the last system. The applied costs for using PSH
positively affected the overall project financing scheme. The cash flow of
the hybrid system with the PSH unit is summarized in Figure A2.
6. Discussion
The study centers on integrating energy storage, specifically Pumped
Storage Hydropower (PSH), into a hybrid renewable energy system to
address the challenges posed by the intermittency of renewable energy
sources [31]. The primary focus is optimizing the operation of PSH units
to achieve cost savings, reduce carbon emissions, and enhance the
overall sustainability of the energy system [10,36].
H.M. Ramos et al.
Fig. 18. Results of grid-in, grid-out, and cost applied within different schemes and seasons.
The preliminary data processing has revealed that the distribution
profile of electricity generation from photovoltaic (PV) and wind turbine
technologies varies across seasons [37]. Solar energy is abundant in
summer due to the high intensity of solar radiation, whereas winter
poses a challenge for meeting demand using wind and solar energy
alone. The daily load pattern exhibits fluctuations, with peak demand
during specific hours attributed to various factors such as industry,
services, air conditioning, heating, and electronic equipment usage.
These findings underscore the need for energy storage systems, as solar
and wind energy alone cannot consistently meet the hourly demand
[18].
To address this energy limitation, the study explores the integration
of PSH into the energy system. PSH provides a means to store excess
energy during high renewable energy generation periods and release it
when renewable energy sources fall short of meeting demand [7]. The
integration of PSH is presented in different schemes, ranging from
Scheme 1, where no energy storage system is used, to Scheme 4, the
optimal solution for integrating PSH.
The study evaluates the annual cost and emissions implications of
these different schemes. It is observed that the schemes incorporating
PSH, especially Scheme 4, result in reduced grid-in costs, leading to cost
savings and a corresponding reduction in carbon emissions. The com­
parison reveals that the contribution of diesel (grid-in) is minimized,
resulting in lower CO₂ emissions. This finding reinforces the
16
Renewable Energy 221 (2024) 119732
environmental benefits of integrating PSH into the system [38].
Schemes 2, 3, and 4 involve more advanced scheduling of PSH op­
erations, considering factors like state of charge, demand, and price
fluctuations. These schemes yield better cost-efficiency and demonstrate
the potential for financial viability. The results show that the integrated
PSH system reduces operational costs and contributes to sustainability
by curbing carbon emissions [38,39]. The favorable net present value
(NPV) and positive cash flow of the hybrid system with PSH reinforce
the financial viability of this integration.
In summary, this study highlights the significance of energy storage
solutions in mitigating the intermittency of renewable energy sources
and enhancing the sustainability of the energy system. Integrating PSH,
particularly in optimized scheduling scenarios, emerges as a promising
strategy for cost reduction, emission reduction, and financial viability
[6]. The findings provide valuable insights for policymakers and
stakeholders in the renewable energy sector seeking to improve the ef­
ficiency and sustainability of energy systems. Further research may
explore real-world implementation and assess such integration’s
broader economic and environmental impacts.
7. Conclusions
Addressing climate change and energy crises is considered one of the
world’s biggest problems in terms of energy availability. Central to
H.M. Ramos et al. Renewable Energy 221 (2024) 119732

Fig. 19. Results of pumping ratio to turbining energy, pumping and turbining within different schemes and seasons.

Fig. 20. Results of CO₂ estimation within different schemes and seasons.

17
H.M. Ramos et al. Renewable Energy 221 (2024) 119732

addressing climate change is proven and reliable energy storage, which
is needed to support intermittent renewables for a rapid energy sector
transition from conventional fossil fuels to decarbonized sources. Pum­
ped storage hydro (PSH) is a fixed storage method that uses a system of
two interconnected water reservoirs, one higher than the other, creating
a different energy potential that can be converted into electrical energy.
For this study, a complete model was developed to simulate different
hybrid systems and improve the differences obtained for each season.
The reported results have confirmed the following main conclusions:
with Scheme 4 in summer, while daily costs can be reduced by more
than 90 % in all seasons. With an optimized and adjustable scheme, PSH
can also provide a broader range of up- and down-ramps while varying
the power system in generation and pumping modes. A hybrid genset
combined with PSH has shifted, stored, and reused generated energy
until an appropriate load is available for system reserves and variable
energy integration.
CRediT authorship contribution statement

• A PSH energy solution can minimize CO2 emissions by reducing the
accumulated grid-in energy supplied by any other fossil fuel-based
generation.
• PSH reduced total cost by optimizing integration with State of
Pumping (SoP) and intermittent renewable sources. The constraints
are met, and objectivity was established to minimize the total cost
considering the PSH scheme.
• The daily integration of the PSH unit into the hybrid mode has
considered the energy sources’ interrelationships. Intermittent
renewable energy sources have different energy patterns than fossil
fuel generation, such as diesel. Focusing on supply and demand while
minimizing costs and reducing the carbon footprint will behave
differently in each situation:
- Different ecological and geographical conditions in different lati­
tudes and longitudes.
- Distinct atmospheric data, i.e., ambient temperature, solar radiation,
wind speed, and other related parameters, contribute to the unique
situation of a hybrid unit.
- Different load and demand patterns and their relative magnitudes to
the other hourly.
It can also be concluded from all the programs that PSH’s contri­
bution is valuable in the sustainable energy transition, as evidenced by
an interesting project financing that shows a positive NPV value and an
interesting payback value of 5 years. CO₂ reduction can be up to 84 %
Helena M. Ramos: Conceptualization, Methodology, Investigation,
Formal analysis, Data curation, Visualization, Writing – original draft,
Writing – review & editing, Resources, Supervision. Jeremy E. Sintong:
Conceptualization, Methodology, Investigation, Formal analysis, Data
curation, Visualization, Writing – original draft, Writing – review &
editing. Alban Kuriqi: Conceptualization, Methodology, Investigation,
Formal analysis, Data curation, Visualization, Writing – original draft,
Writing – review & editing.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgments
Helena M. Ramos and Alban Kuriqi are grateful for the support of the
Foundation for Science and Technology through funding UIDB/04625/
2020 of the CERIS research unit.

Nomenclature

Abbreviations and Acronyms

CAES Compressed-Air Energy Storage
CO₂ Carbon Dioxide
DSM Demand Side Management
EUR Euro currency
FSL Full Supply Level
kW Kilo Watt
kWh Kilo Watt-Hour
LFP Lithium-Ion Battery
LHV Lower Heating Value
MiOL Minimum Operating Level
MWh Megawatt-Hour
PSH Pumped Storage Hydropower
PV Solar Photovoltaic
SoP State of Pumping
SoPInit Initial State of Pumping
UR Upper Reservoir
VRE Variable Renewable Energies
VRF Vanadium Redox Flow Battery
WT Wind Turbine
Wh Watt-Hour

18
H.M. Ramos et al. Renewable Energy 221 (2024) 119732

◦
C Degree Celsius
m3 Cubic-Meter
g Gravitational acceleration (9.8 m/s2)
kg Kilogram

Symbols
PPV (t) Power Output from Solar PV on an hourly basis
PRated PV Rated Power of Solar PV
GT (t) Hourly Irradiance rate
Gref Irradiance Reference of Solar PV
kT Temperature coefficient of Solar PV
TC (t) Cell’s temperature of Solar PV on an hourly basis
Tref Assumed reference temperature of Solar PV during test conditions
Tamb (t) Ambient Temperature on an hourly basis
PWT (t) Power Output from Wind Turbine on an hourly basis
ηWT The efficiency of Wind Turbine
Cp Power Coefficient of Wind Turbine
R Rotor Radius of Wind Turbine Blade
U(t) Hourly Wind Speed
UCut− In Cut-In Wind Speed applied in Wind Turbine
Urated Rated or Nominal Wind Speed applied in Wind Turbine
UCut− Out Cut-Out Wind Speed applied in Wind Turbine
ρa (t) Air Density on an hourly basis
Celec.price (t) Electricity Price on an hourly basis
PDemand or Load (t) Demand or Load Power on an hourly basis
Pturbining (t) Power extracted during turbine mode on an hourly basis
Ppumping (t) Power absorbed during pumping mode on an hourly basis
Pmax.turbining (t) Maximum turbine power on an hourly basis
Pmax.pumping (t) Maximum pumping power on an hourly basis
PGrid− In (t) Power injected from the Grid into Hybrid System on an hourly basis
PGrid− Out (t) Power supplied to the Grid from Hybrid System on an hourly basis
PSupply (t) Power supplied from renewable hybrid generation on an hourly basis
SOPmax
(t) Maximum State of Pumping on an hourly basis
VRmin Minimum Reservoir Volume
VRmax Maximum Reservoir Volume
VR (t) Reservoir Volume on an hourly basis
ηpumping Efficiency in pumping mode
ηturbining Efficiency in turbine mode
ρdiesel Assumed Density of Diesel as fuel for generation
LHVdiesel Assumed Lower Heating Value of Diesel Fuel
Vpumping (t) Volumetric measurement in pumping mode on an hourly basis
Vturbining (t) Volumetric measurement in turbine mode on an hourly basis
ERCO2 Assumed Emissions Rate of CO₂ produced from diesel generation

Appendix A

19
H.M. Ramos et al. Renewable Energy 221 (2024) 119732

Fig. A1. Preliminary load and renewables supplied generation profile in four seasons.

20
 H.M. Ramos et al.
21

Fig. A2. Project Cash Flow of the Integration of PSH solution with Intermittent Renewables Scheme.

Renewable Energy 221 (2024) 119732

H.M. Ramos et al. Renewable Energy 221 (2024) 119732

Fig. A3. Daily Pumping and Turbining Volumetric Profile within Different Schemes and Seasons.

22
H.M. Ramos et al. Renewable Energy 221 (2024) 119732

Fig. A4. Daily Pumping and Turbining Energy Distribution Profile within Different Schemes and Seasons.

23
H.M. Ramos et al. Renewable Energy 221 (2024) 119732

References [21] R. Uria-Martinez, M.M. Johnson, R. Shan, N.M. Samu, G. Oladosu, J.M. Werble,
H. Battey, US Hydropower Market Report, Oak Ridge National Lab.(ORNL), Oak
Ridge, TN (United States), 2021.
[1] S. Koohi-Fayegh, M.A. Rosen, A review of energy storage types, applications and
[22] M. Lafratta, M. Leach, R.B. Thorpe, M. Willcocks, E. Germain, S.K. Ouki, A. Shana,
recent developments, J. Energy Storage 27 (2020), 101047.
J. Lee, Economic and carbon costs of electricity balancing services: the need for
[2] T. Brijs, A. Belderbos, K. Kessels, D. Six, R. Belmans, F. Geth, Energy storage
secure flexible low-carbon generation, Energies (2021).
participation in electricity markets, Advan. Energy Storage (2022) 775–794.
[23] IHA, Innovative Pumped Storage Hydropower Configurations and Uses,
[3] M.S. Javed, T. Ma, J. Jurasz, M.Y. Amin, Solar and wind power generation systems
International Hydropower Association (IHA), London, United Kingdom, 2021,
with pumped hydro storage: review and future perspectives, Renew. Energy 148
pp. 3–93.
(2020) 176–192.
[24] T. Mindra, G. Florea, O. Chenaru, R. Dobrescu, L. Toma, Combined peak shaving/
[4] P. Gajewski, K. Pieńkowski, Control of the hybrid renewable energy system with
time shifting strategy for microgrid controlled renewable energy efficiency
wind turbine, photovoltaic panels and battery energy storage, Energies (2021).
optimization, in: 2022 21st International Symposium INFOTEH-JAHORINA
[5] A. Altan, S. Karasu, E. Zio, A new hybrid model for wind speed forecasting
(INFOTEH), 2022, pp. 1–6.
combining long short-term memory neural network, decomposition methods and
[25] M. Liebensteiner, A. Haxhimusa, F. Naumann, Subsidized renewables’ adverse
grey wolf optimizer, Appl. Soft Comput. 100 (2021), 106996.
effect on energy storage and carbon pricing as a potential remedy, Renew. Sustain.
[6] J.-F. Zhao, U.-J. Oh, J.-C. Park, E.S. Park, H.-B. Im, K.Y. Lee, J.-S. Choi, A review of
Energy Rev. 171 (2023), 112990.
world-wide advanced pumped storage hydropower technologies, IFAC-
[26] M. Stocks, R. Stocks, B. Lu, C. Cheng, A. Blakers, Global atlas of closed-loop
PapersOnLine 55 (9) (2022) 170–174.
pumped hydro energy storage, Joule 5 (1) (2021) 270–284.
[7] K.R. Vasudevan, V.K. Ramachandaramurthy, G. Venugopal, J.B. Ekanayake, S.
[27] J.W. Saulsbury, A Comparison of the Environmental Effects of Open-Loop and
K. Tiong, Variable speed pumped hydro storage: a review of converters, controls
Closed-Loop Pumped Storage Hydropower, Pacific Northwest National Lab.
and energy management strategies, Renew. Sustain. Energy Rev. 135 (2021),
(PNNL), 2020. Richland, WA (United States).
110156.
[28] REN, Market Load Information in MWh, 2019. https://mercado.ren.pt/EN/Elec
[8] H. Ye, L. Zhang, M. Yang, G. Liang, S. Liu, G. Li, W. Tsang, H. Xie, Variable-speed
tr/SystemsManagement/Load/Pages/Actual.aspx. (Accessed 27 May 2022).
pumped hydro storage technology: overview, solutions and case studies, in: 2021
[29] K. Mohammadi, O. Alavi, A. Mostafaeipour, N. Goudarzi, M. Jalilvand, Assessing
6th International Conference on Power and Renewable Energy (ICPRE), 2021,
different parameters estimation methods of Weibull distribution to compute wind
pp. 1273–1278.
power density, Energy Convers. Manag. 108 (2016) 322–335.
[9] M.B. Mustafa, P. Keatley, Y. Huang, O. Agbonaye, O.O. Ademulegun, N. Hewitt,
[30] R. Winkel, U. Weddige, D. Johnsen, V. Hoen, S. Papaefthimiou, Shore side
Evaluation of a battery energy storage system in hospitals for arbitrage and
electricity in Europe: potential and environmental benefits, Energy Pol. 88 (2016)
ancillary services, J. Energy Storage 43 (2021), 103183.
584–593.
[10] J. Tian, L. Yu, R. Xue, S. Zhuang, Y. Shan, Global low-carbon energy transition in
[31] J. Jurasz, F.A. Canales, A. Kies, M. Guezgouz, A. Beluco, A review on the
the post-COVID-19 era, Appl. Energy 307 (2022), 118205.
complementarity of renewable energy sources: concept, metrics, application and
[11] A. Dutta, Energy conservation and its impact on climate change, in: P.K. Sikdar
future research directions, Sol. Energy 195 (2020) 703–724.
(Ed.), Environmental Management: Issues and Concerns in Developing Countries,
[32] J. Almorox, C. Voyant, N. Bailek, A. Kuriqi, J.A. Arnaldo, Total solar irradiance’s
Springer International Publishing, Cham, 2021, pp. 139–150.
effect on the performance of empirical models for estimating global solar radiation:
[12] IRENA, Global renewables outlook, Energy Transformation 2050, International
an empirical-based review, Energy 236 (2021), 121486.
Renewable Energy Agency Abu Dhabi, 2020.
[33] L. Chen, Q. Xu, Y. Yang, J. Song, Optimal energy management of smart building for
[13] IEA, Energy transitions, Tracking Progress in Clean Energy Transitions through Key
peak shaving considering multi-energy flexibility measures, Energy Build. 241
Indicators across Fuels and Technologies, 2021.
(2021), 110932.
[14] H. Liu, I. Khan, A. Zakari, M. Alharthi, Roles of trilemma in the world energy sector
[34] H.M. Ramos, B. Vargas, J.R. Saldanha, New integrated energy solution idealization:
and transition towards sustainable energy: a study of economic growth and the
hybrid for renewable energy network (Hy4REN), Energies (2022).
environment, Energy Pol. 170 (2022), 113238.
[35] M. Guezgouz, J. Jurasz, B. Bekkouche, T. Ma, M.S. Javed, A. Kies, Optimal hybrid
[15] O. Schmidt, S. Melchior, A. Hawkes, I. Staffell, Projecting the future levelized cost
pumped hydro-battery storage scheme for off-grid renewable energy systems,
of electricity storage technologies, Joule 3 (1) (2019) 81–100.
Energy Convers. Manag. 199 (2019), 112046.
[16] IRENA, Global renewables outlook, Energy Transformation 2050, International
[36] IEA, Energy transitions, Tracking Progress in Clean Energy Transitions through Key
Renewable Energy Agency Abu Dhabi, 2021.
Indicators across Fuels and Technologies, 2020.
[17] K.M. Tan, T.S. Babu, V.K. Ramachandaramurthy, P. Kasinathan, S.G. Solanki, S.
[37] A. Ouammi, H. Dagdougui, R. Sacile, A. Mimet, Monthly and seasonal assessment
K. Raveendran, Empowering smart grid: a comprehensive review of energy storage
of wind energy characteristics at four monitored locations in Liguria region (Italy),
technology and application with renewable energy integration, J. Energy Storage
Renew. Sustain. Energy Rev. 14 (7) (2010) 1959–1968.
39 (2021), 102591.
[38] A. Botterud, T. Levin, V. Koritarov, Pumped Storage Hydropower: Benefits for Grid
[18] L. Malka, A. Daci, A. Kuriqi, P. Bartocci, E. Rrapaj, Energy Storage Benefits
Reliability and Integration of Variable Renewable Energy, Argonne National Lab.
Assessment Using Multiple-Choice Criteria: the Case of Drini River Cascade,
(ANL), Argonne, IL (United States), 2014.
Energies, Albania, 2022.
[39] G. Krajačić, N. Duić, Z. Zmijarević, B.V. Mathiesen, A.A. Vučinić, M. da Graça
[19] E. López González, F. Isorna Llerena, M. Silva Pérez, F. Rosa Iglesias, J. Guerra
Carvalho, Planning for a 100% independent energy system based on smart energy
Macho, Energy evaluation of a solar hydrogen storage facility: comparison with
storage for integration of renewables and CO2 emissions reduction, Appl. Therm.
other electrical energy storage technologies, Int. J. Hydrogen Energy 40 (15)
Eng. 31 (13) (2011) 2073–2083.
(2015) 5518–5525.
[20] K. Mongird, V. Viswanathan, J. Alam, C. Vartanian, V. Sprenkle, R.J.E. Baxter,
2020 Grid Energy Storage Technology Cost and Performance Assessment, vol.
2020, 2020, pp. 6–15.

24