Offshore Wind Power Integration into Future Power Systems: Overview and Trends

Nowadays, wind is considered as a remarkable renewable energy source to be implemented in power systems. Most wind power plant experiences have been based on onshore installations, as they are considered as a mature technological solution by the electricity sector. However, future power scenarios and roadmaps promote offshore power plants as an alternative and additional power generation source, especially in some regions such as the North and Baltic seas. According to this framework, the present paper discusses and reviews trends and perspectives of offshore wind power plants for massive offshore wind power integration into future power systems. Different offshore trends, including turbine capacity, wind power plant capacity as well as water depth and distance from the shore, are discussed. In addition, electrical transmission high voltage alternating current (HVAC) and high voltage direct current (HVDC) solutions are described by considering the advantages and technical limitations of these alternatives. Several future advancements focused on increasing the offshore wind energy capacity currently under analysis are also included in the paper.


Introduction
Energy demand has been increasing non-stop during the last decades [1]. Nowadays, fossil fuel sources (i.e., coal, oil and natural gas) provide around 85% of the world energy demand, according to the BP Energy Outlook of 2019 [2]. However, with the Paris climate agreement established in December 2015, this energy scenario is about to change [3]. This climate agreement aims to restrict maximum increase in the global average temperature below 2 • C above pre-industrial levels [4]. To fulfill this goal, greenhouse gas (GHG) emission trends should drastically change [5]. Consequently, the use of fossil fuels should be reduced, as they are considered as the main source of GHG emissions [6]. Actually, global GHG emissions are dominated by the emissions of CO 2 due to the combustion of fossil fuels, which has been increasing continuously since 1990 [7]. The power sector should be decarbonized by 2050 to meet the Paris agreement target [8]. Furthermore, Liddle and Sadorsky estimated that increasing by 1% the share of non-fossil fuel electricity generation can reduce by up to 0.82% the CO 2 emissions [9]. This environmental worry is one of the reasons to promote the integration of renewable energy sources (RES) into power systems [10]. Moreover, RES can also mitigate the energy dependence on fossil fuels imported from other countries [11]. Apart from the economic costs of these fossil fuel imports, decreasing energy dependence increases electricity supply security [12]. The International Energy Agency defines electricity supply security as the uninterrupted availability of energy sources at an affordable price [13]. However, political stability, market liberalization and foreign affairs are nowadays linked to energy supply security [14]. As a consequence, it is important to be energy-independent to guarantee the energy security of a country [15].
While RES provide an acceptable solution for these two problems, they also face many challenges as their integration increases into the grids, mostly based on their intermittency, variability and uncertainty due to their dependency on weather conditions [16]. Actually, they are usually considered as 'non-dispatchable' sources [17]. This fact makes them hard to integrate into power systems [18], as transmission system operators (TSOs) have to deal with not only the uncontrollable demand but also uncontrollable generation [19,20]. RES include bioenergy, geothermal energy, hydropower, ocean energy (tide and wave), PV, thermal solar energy and wind energy (onshore and offshore) [21]. Some of them (such as wind and solar installations) are connected to the grid through power electronic converters, reducing the rotational inertia of the system as they replace conventional generation units [22,23]. This fact compromises the frequency stability and alters the transient response [24]. As a result, several frequency control strategies have been proposed in the specific literature [25][26][27][28][29][30]. Other alternatives to increase the RES share in power systems and avoid the aforementioned problems are to complement one source with another (for instance, wind with solar and/or hydropower) [31][32][33] or to use storage systems (such as flywheels, pumped hydroelectric storage, batteries, hydrogen, etc.) [34,35].
Among these renewable technologies, wind is one of the most economic, prominent and matured RES technologies [36,37]. In fact, since 2001, global cumulative installed wind capacity has shown an exponential growth, as can be seen in Figure 1a. Among the total wind capacity, 23 GW came from offshore installations in 2018, compared to 1 GW in 2007, refer to Figure 1b [38]. Despite offshore wind energy dating back to the 1990s, its popularity started around ten years ago [39]. This increase is due to the current interest of the wind energy industry in offshore wind power [40]. For instance, offshore wind energy investments surpassed onshore investments in Europe in 2016, as presented in [41]. Moreover, nearly 40% of the total wind capacity is expected to come from offshore wind energy in Europe in 2030 [42,43].
In addition, offshore wind energy presents many advantages compared to onshore wind power plants, especially related to wind energy potential [44,45]: (i) Offshore mean wind speeds are higher and wind power variability is also lower than onshore wind power; (ii) their visual and acoustic impact is usually lower than onshore; subsequently (iii) larger wind turbines (WTs) can be installed [46]. Actually, on the European coasts, the available offshore wind energy is about 350 GW [47]; the USA's shores present an offshore wind power potential of more than 2000 GW [48]; the offshore wind resource in China is about 500 GW in water depth under 50 m [49]; and the east and west Indian coasts have an offshore wind potential of 4.4 GW and 6.7 GW, respectively [50].
Furthermore, offshore wind speed usually increases with distance from the shore, thus increasing the power generated, as it depends on the cube of the wind speed [51]. However, higher installation and maintenance costs of offshore wind power plants (OWPP) far from the shore balance the benefits of higher energy production [52]. Indeed, OWPP are around 50% more expensive than onshore wind power plants [53], but their costs are expected to decline up to 35% by 2025 [54]. The global weighted average levelized cost of energy (LCOE) in 2018 was 20% lower than in 2010. These cost reductions can be a result of [55]:

•
The evolution in wind turbine technology, installation and logistics • The economies of scale in operations and maintenance • The improved capacity factors due to higher hub heights, better wind resources and larger rotor diameters This paper analyzes and reviews different aspects of offshore wind power plants, including several future alternatives to increase the offshore wind power capacity. The rest of the paper is organized as follows: Section 2 presents the current status of offshore wind power plants (WTs and OWPP sizes, water depth, distance from shore and electrical transmission to shore). Future advancements possible for larger offshore wind power plant integration are analyzed in Section 3. Finally, Section 4 gives the conclusions.

Preliminaries: Classification of Wind Turbines
WTs are usually classified as fixed speed wind turbines (FSWTs) and variable speed wind turbines (VSWTs) [56]. FSWTs work at the same rotational speed regardless of the wind speed [57]. VSWTs can operate around their optimum power point for each wind speed, using a partial or full additional power converter [58]. As a result, VSWTs are more efficient than FSWTs [59]. Moreover, WTs present different topologies depending on their generator [60]: type 1 includes a squirrel cage induction generator; type 2 includes a wound rotor induction generator; type 3 includes a doubly-fed induction generator (DFIG); and type 4 includes a full-converter synchronous generator [61]. Types 1 and 2 are FSWTs, whereas types 3 and 4 are VSWTs.

•
The DFIG configuration needs a gearbox, generator and partial-scale power converter (around 30%), as shown in Figure 2a. The gearbox couples the blades with the generator, increasing the rotational speed from the rotor hub to the induction machine [73][74][75]. The stator is directly connected to the grid, whereas the rotor is connected to the power converter [76]. As a result, the converter only covers the power produced by the rotor of the DFIG [77].

•
The synchronous generator of a full-converter WT is excited by an external DC source or by permanent magnets [78]. In this case, the hole generator is connected to the grid through a power converter [79]. Hence, all the generated power from a WT can be regulated accordingly [80]. They have low maintenance costs and negligible rotor losses [81]. Moreover, some type 4 WTs have no gearbox, as depicted with a dotted line in Figure 2b, using a direct driven multipole generator [82]. In Europe, the rated capacity of offshore WTs has been continuously increasing during the last decade. For instance, in 2017, the average rated capacity of WTs was 5.9 MW, compared to 3 MW in 2010 and 4.8 MW in 2016 [83]. In 2018, new offshore WTs were 6.8 MW on average, 15% larger than in 2017. Comparing 2018 to 2010, the average WT increase is more than 200%. Moreover, two 8.8 MW offshore WTs were installed in the United Kingdom in 2018, those being the largest WTs installed of the world. The commercial model of those WTs was V164-8.8 MW from MHI Vestas Offshore [84]. However, nowadays there are larger commercial offshore WTs, up to 12 MW, as presented in [85]. Table 1 shows the 10 largest WTs currently available. All of them have a rated power over 8 MW, rotor diameters between 150 and 200 m and are equipped with synchronous generators (type 4). Regarding OWPP capacity, it has also increased dramatically in the last 10 years (around 700%), in line with the increase of average offshore WT capacity. In fact, average OWPP capacity was 79.6 MW in 2007. In contrast, 561 MW was the average capacity for OWPPs in 2018 [84]. This means that considering the average WT and OWPP capacities of the year 2018, each OWPP has between 80 and 85 WTs. On the other hand, OWPP depth and distance to the shore have not increased that much in recent years. At the end of 2013, the average water depth of OWPPs was 16 m with an average distance to the shore of 29 km [86]. In 2018, the average water depth of OWPPs under construction was 27.1 m, with an average distance to shore of 33 km. This means that water depth has increased by 170% and distance to shore by around 110%. There are some OWPPs that should be mentioned: Hornsea One (UK) and EnBW Hohe See (Germany) are the OWPPs located farthest from the shore (103 km away); Kincardine Pilot (Scotland), a floating demonstration project, has a water depth of 77 m [84]; and Hywind (Scotland), the first fully operational floating wind farm, with water depths varying between 95 and 129 m [87].

Offshore Wind Power Electrical Power Transmission
For the electrical power transmission from the OWPP plant to the shore, there are two possibilities: (i) High voltage alternating current (HVAC) and (ii) high voltage direct current (HVDC). Figure 3 depicts an overview of the current state of offshore wind power energy transmission to the shore.  [88]. Their easy protection system design and the use of transformers to change between different voltage levels were the main reasons to use them [89]. However, the high capacitance of submarine HVAC cables combined with the low resistivity of sea water caused different electromagnetic dynamic and transient problems from those of conventional overhead lines, such as distortion of the voltage's shape due to resonance problems [90,91]. This high capacitance also leads to substantial charging currents, subsequently reducing the active power transmission capacity and transmitting reactive power in long distances [92]. A possible solution could be to install reactive power compensation units along the HVAC submarine cables, but they are expensive devices and it is a difficult task to carry out [93]. an alternative found in the literature is to add compensation units only at both ends (onshore and offshore) of the underwater cables, which improve the current profile along them [94,95]. However, their effect is very limited for distances over 60-75 km [96,97]. With the aforementioned considerations, the topology for HVAC transmission from OWPPs is depicted in Figure 4 [98,99]. It consists of:

•
An offshore substation to increase the offshore voltage level (usually from 30-36 kV) to the transmission voltage level at 132-400 kV. • Three-core HVAC submarine transmission cables.

•
Reactive compensation units on both ends (offshore and onshore), such as static VAR compensators (SVCs) or static synchronous compensators (STATCOMs).

•
An onshore substation, if the onshore interconnecting grid voltage is different from the offshore transmission system rated voltage. As can be seen, the OWPP grid is synchronously coupled to the main onshore grid. This is another problem of HVAC links for OWPPs, as all faults in either the grid or the OWPP are propagated to the other one [100].

High Voltage DC
High voltage direct current (HVDC) transmission is considered as the best solution to OWPPs located far away from the land [101]. Actually, some studies conclude that HVDC links are economically viable for distances above 50-70 km [102]. A graphical comparison of costs between HVAC and HVDC transmission systems can be seen in Figure 5 Figure 6 shows the main elements of an HVDC connection, and consists of [104,105]:  Figure 6. Offshore wind power plant HVDC transmission system.
In contrast to the HVAC topology presented in Figure 4, the HVDC link electrically decouples both the OWPP and the onshore grid, avoiding the propagation of possible disturbances between them [106,107].
Two different HVDC technologies are currently under use: Line-commutated converters (LCCs) based on thyristors and voltage-source converters (VSCs) based on insulated gate bipolar transistors (IGBTs) [108]. Among them, there is not a clear consensus about which technology is better: Some authors consider that LCCs are superior to VSCs in terms of reliability, cost and efficiency [26], whereas others affirm that the VCS-HVDC transmission system is the most promising technology [109]. a comparison between both HVDC technologies is summarized in Table 2 [110]. Recently, another technology called the diode rectifier unit (DRU) has been under discussion, though has not been implemented yet [111].

Line-Commutated Converters
Traditionally, HVDC transmission systems have been based on LCCs, which use thyristors as the base technology. Actually, LCC is a trusted and mature technology [112] that links a mainland with some islands (e.g., in Northern Europe [113]). However, solutions based on thyristors usually involve the injection of some harmonics. For example, a twelve-pulse thyristor bridge, which is made up of two six-pulse bridges; the fifth and seventh harmonics can be canceled [114]. The LCC-HVDC transmission system is based on this twelve-step bridge [115,116].
The main drawbacks of the LCC-HVDC link are [116][117][118][119][120]: • It can only transfer power between (at least) two active grids. As a result, an auxiliary start-up system is necessary in the OWPP.

•
It demands reactive power, which needs to be supplied through reactive support devices.

•
Despite most harmonics being canceled by using a twelve-pulse bridge, others still remain, thus needing additional filters. • It requires voltage support for the OWPP AC bus. Two possible solutions can be found to overcome this requirement: (i) Installing a dedicated STATCOM, which increases considerably the overall cost or (ii) controlling the turbine inverters individually, which is technically challenging.

•
The inverter is susceptible to commutation failures, especially when connected to weak AC power systems.
Blasco et al. suggest that the filter's design depends on the harmonic characteristics of the AC grid and the active power exchanged by the LCC-HVDC link, thus needing a detailed AC power system analysis [121].

Voltage-Source Converters
Since 2005, VSC-HVDC technology has been used in offshore applications. It is based on IGBTs [122]. The main characteristics of the VSC-HVDC link are summarized as [123][124][125]: • It can control active and reactive power simultaneously.

•
It can feed island-mode, weak AC and passive networks. • Its station requires less space than that of an LCC (about 60% less).

•
The cables are lighter. • It does not require reactive power compensation.

•
It can transmit power from zero to full-rating bidirectional, enabling OWPP start-up (black start operation) and working at low wind speeds.
Despite all these advantages, VSC-HVDC presents higher commutation losses and costs compared to the LCC-HVDC. Moreover, it can only handle limited voltage and power levels [126].

Diode Rectifier Unit
During the last years, DRU-HVDC has been under discussion. A DRU includes several diodes, a transformer and a smoothing reactor [127]. As DRU can only convert AC to DC [128], a hybrid topology combining DRU and VSC/LCC must be used, introducing the DRU as the offshore rectifier and the LCC/VSC as the onshore converter [129]. The main advantages of DRU-HVDC compared to LCC-HVDC and VSC-HVDC are [130][131][132][133]: • Reduction of volume (80%) and weight (66%) of the platform. However, several problems have to be solved before implementing a DRU-HVDC connection [132][133][134][135]: • As the DRU is a non-controllable passive device, the OWPP AC system must be regulated and controlled by the WT, thus requiring different WT and OWPP schemes.

•
The onshore converter (LCC/VSC) controls the HVDC voltage. Subsequently, the DRU output DC voltage must be higher than the minimum voltage value to start conducting and transmit the power to the onshore station.

•
Passive filters or active compensation devices are needed to remove the harmonic currents injected by the DRU.

•
Voltage and frequency control stages are needed in the offshore grid for DRU commutation. • A DRU is not able to provide auxiliary active power for the WT and OWPP substation, being then a drawback to the self-start of WTs. • A DRU is not able to provide reactive power, needing power converters or other devices to compensate it.

Potential Future Advancements for Offshore Wind Energy
In 2018, ENTSO-E and ENTSO-G published their Ten Year Network Development Plan (TYNDP) scenarios. It was the first time that both European electrical and gas TSOs collaborated together. The TYNDP 2018 covers from 2020 to 2040 [136]. In 2030-2040, it is expected that between 45 and 75% of the overall European demand will be covered by RES, especially by hydro, wind and solar power. Actually, in the North Sea and Baltic Sea regions, the offshore wind power capacity is estimated to reach between 40 and 59 GW in 2030, and between 86 and 127 GW in 2040, according to the TYNDP. Other authors propose similar offshore wind power capacity scenarios in these regions; for instance, scenarios were modeled and optimized by Koivisto and Gea-Bermudez [137]. Greenpeace published in 2015 their 'Energy [R]evolution' forecast, where 148 GW of offshore wind capacity is expected to be installed in Europe in 2050 [138]. In the US, the Department of Energy considers that 22 GW of OWPPs can be installed by 2030, increasing up to 86 GW by 2050 [139]; according to the scenarios presented by the Energy Resources Institute of India for 2050, it could have 170 GW installed of offshore wind energy by then [140]; and the Chinese scenarios propose to install 200 GW of OWPPs (150 GW near offshore wind and 50 GW far offshore wind) by 2050 [141,142]. By these means, OWPPs seem to have an important energy role in the future worldwide.
However, onshore wind and other conventional fossil fuel technologies are currently cheaper than offshore wind energy [143]. As a consequence, different alternatives are being researched to reduce further costs of offshore wind power development: • Power-to-X conversion (P2X) • North Sea Wind Power Hub: The Hub-and-Spoke project • Offshore storage options These initiatives are discussed in detail in the following.

Power-to-X Conversion
P2X is based on converting power (electricity) to diverse substances (X) [144]. The different alternatives available in the P2X conversion are [145,146]: • Power-to-heat (P2H): The electrical generation excess is linked to a heat device (electric boiler, heat pump), avoiding any intermediate energy carrier and subsequently increasing the global efficiency. • Power-to-liquid (P2L): Different alternatives can be found in the specific literature, including the production of syngas through hydrogenation of CO 2 and reverse water gas shift; co-electrolysis of CO 2 and H 2 O; or directly through the electro-reduction of CO 2 to methanol. • Power-to-chemicals (P2C): From the syngas obtained with the power-to-liquid conversion, several compounds can be produced accordingly.

•
Power-to-gas (P2G): Hydrogen is obtained from an electrolysis process and the possible subsequent conversion to methane with CO 2 . • Power-to-mobility (P2M): The electrical generation excess is used by the mobility sector through electric vehicles with an electric motor of 90% efficiency instead of an internal combustion engine (efficiency of 20%) or fuel cell (efficiency of 50%).

•
Power-to-power (P2P): Electricity is converted into chemical or mechanical energy, which is stored and later reconverted into electric power.
These transformations are expected to be very relevant in future power systems, as the generated electricity excess can be stored in different ways and later used as, for instance, fuel for power plants [147]. Hence, the system's flexibility would be enhanced [148]. By these means, high capacity P2X plants could increase the RES supply by providing supply security in terms of storage facilities [149]. Moreover, as explained in [150], the P2X conversion provides a real link between different sectors, promoting the transition towards a future urban smart energy system. As an example, Figure 7 depicts the power-to-heat conversion joint [151]. an exhaustive analysis of 128 P2X demo projects in operation in Europe is discussed in [152]. These projects aim to gain experience with system integration of P2X components. Moreover, Denmark is interested in the conversion of electricity to hydrogen and liquid fuels through P2X solutions. Thus, they can become a front-runner in this technology as large Danish companies already work with it [153].  Several authors have already analyzed these technologies combined with the wind resource. Different flexibility options for wind power plants are analyzed in [154], concluding that the P2H solutions provide the most cost-effective scenarios with the lowest CO 2 emissions. Pursiheimo et al. focused on the feasibility of the P2G technology in Nordic countries to achieve a 100% RES system. The main applications of P2G are focused on supplying gas to transport and industrial sectors [155]. Furthermore, the use of P2G has been proved in Denmark to be a successful tool to complement wind power plants [156]. However, both investment costs of facilities and energy losses (due to the low efficiency in the conversion process) are high. Hence, the hydrogen produced from wind power also has a high cost [157]. For instance, in [158] different energy applications of hydrogen (P2P, P2G, P2M) are considered for a hybrid offshore wind-hydrogen power plant in France, obtaining negative profits due to the high investment costs in both wind and hydrogen infrastructures. Other authors conclude that the combination of a wind-hydrogen power plant should be considered to sell hydrogen directly, as re-powering hydrogen for electricity is extremely expensive [159]. Consequently, future works should be focused on reducing investment and maintenance costs for such power conversion solutions.

North Sea Wind Power Hub: The Hub-and-Spoke Project
To meet the Paris Agreement and the GHG reduction goals (refer to Section 1) in the countries around the North Sea, the North Sea Wind Power Hub (NSWPH) consortium was created. TenneT (a Dutch-German electricity TSO), Port of Rotterdam (the biggest port in Europe), Energinet (a Danish TSO) and Gasunie (a European energy infrastructure company) are the partners of the consortium [160]. NSWPH aims to facilitate the deployment of large scale OWPPs in the North Sea, evaluating and developing the Hub-and-Spoke project. The project consists on several central platforms, called hubs, which are in charge of supporting the power transport infrastructure by using the P2X conversion instead of the offshore converter platforms used currently. According to TenneT, the offshore wind power capacities will be in the range of 70 to 150 GW by the year 2040 and up to 180 GW by 2045 [161], similar values to those proposed in the TYNDP. Two main challenges are identified in NSWPH, (i) a strong power transmission infrastructure and (ii) high flexibility requirements. Mainly due to onshore surface constraints, onshore wind power plants and PV installations are not enough to decarbonize the power systems of this area [162]. The hub-and-spoke concept proposed by NSWPH is made up of several modular hubs located in different zones of the North Sea, which connect OWPPs with bordering North Sea countries. This can be seen in the figure of page 6 of [163]. By using high capacity DC cables, the power generated by OWPPs is transmitted to onshore grids in different locations connected in a smart and coordinated manner. These DC connections also provide high interconnection capacity among the different countries. Moreover, the hub-and-spoke concept can promote onshore OWPP integration through P2G transformation. Power systems thus become flexible through such P2X conversion [163].
Apart from defining the hub-and-spoke project, the NSWPH consortium also aims to demonstrate the technical feasibility of the project. So far, the consortium have concluded that [164]: • The optimal capacity of the OWPP is estimated to be between 10 and 15 GW.

•
Hub substructures can be based on four different foundation types: Ciasson island, sand island, platform and gravity-based structure. A comparison among them is presented in Table 3, as presented in [164].

•
Both the spatial requirements and investment costs of the hubs are similar regardless of being all-electric, all-hydrogen or combining electricity and hydrogen: -All-electric hub-and-spoke: The electricity generated by the OWPP is transmitted to the shore.

-
All-hydrogen hub-and-spoke: The electricity generated by the OWPP is transformed offshore into hydrogen, and transported through pipelines to the shore.

-
Combined electricity and hydrogen hub-and-spoke: combines the two previous concepts. Moreover, the lifetime savings between CAPEX (capital expenditure) and OPEX (operational expenditures) for a 12 GW hub-and-spoke project (Denmark (2 GW), Germany (6 GW) and the Netherlands (4 GW)) could rise up to 2.5 billion e, without considering the P2X conversion, compared to a radial approach. This reduction is due to the lack of additional interconnection capacity between those countries. A study compared the LCOE between hub-and-spoke projects and radial approaches, concluding that the LCOE was able to be reduced for hub sizes between 6 and 12 GW, but limited for capacity hubs between 24 and 36 GW. Furthermore, electricity prices and emissions were also reduced. The total cost saving of a hub-and-spoke project compared to a no-hub project was then estimated to be between 15 and 20 billion e [165]. The main drawback of the hub-and-spoke project is that it would take more than 10 years of development and construction to become operational. Moreover, policies, regulatory framework and market design should be reconsidered to ensure a stable market. As the Paris Agreement must be fulfilled by 2050, these issues should be urgently reconsidered in order to carry out multiple hub-and-spoke projects by then [166].

Offshore Storage Options: Hydrogen and Compressed-Air Energy Storage
As electrical generation has to be immediately sold to supply the electrical consumption, and due to the stochastic nature of RES, energy storage emerges as an important solution for these sources [167]. However, as traditional energy storage technologies are difficult to use in a marine environment, new alternatives are being developed to store offshore energy. Wang et al. provide a comprehensive review on existing marine renewable energy storage solutions [168].

Hydrogen Energy Storage
The surplus of electricity produced by OWPP can be stored as hydrogen, and used later to generate power in fuel cells or as fuel in hydrogen vehicles [169]. Most alternatives available are based on the P2X technology, as previously described in Section 3.1.
An example of such alternatives can be found in the Deep Purple project which is based on the important CO 2 emissions from Norwegian oil and gas production. The project involves TechnipFMC, SINTEF, Subsea Valley and Maritim Forening Sogn og Fjordane, which develop the concept and new technology. It has received funding from the Research Council of Norway [170]. The Deep Purple project aims to convert electricity from OWPP to hydrogen and store such energy on the seabed. The hydrogen can then be used for several purposes [171,172]: • Supply stable and renewable power to oil and gas installations • Supply stable and renewable power to remote islands • Provide a coastal hydrogen infrastructure to maritime sector • Provide local production of power, hydrogen and oxygen to fist farming It is expected to have a full-scale pilot by 2025 in Norway [173]. Figure 8 depicts an overview of the Deep Purple project.

Compressed Air Energy Storage
Compressed air energy storage (CAES) systems are a solution to energy storage based on the compression of air [174]. According to [175], the integration of CAES with wind and solar power generation can increase the RES share rate, as CAES is reported as less expensive than other storage systems, and to be large and powerful enough to store energy on a utility scale level [176]. Moreover, due to the high installation and capital cost of undersea transmission cables, offshore CAES can increase the cable's capacity factor, potentially lowering the average cost of offshore wind power while increasing the reliability and economic value of delivered power [177].
The TAKEOFF Business Incubator (University of Malta) has already patented the a storage technology called FLASC, with the aim of integrating large-scale energy storage into OWPPs. This solution is tailor-made for the offshore market, exploiting existing infrastructure and supply-chains, see Figure 9 [178]. FLASC uses compressed air for energy storage purposes, relying on the hydrostatic pressure of the deep-sea areas to maintain a stable pressure in the compressed air storage. As it uses existing infrastructure, it is considered a cost-effective solution. In [179,180], the multi-system integration and the working principle of the FLASC storage technology are described. This solution can also be used in order to: (i) Convert the intermittent RES supply into a stepped out one, simplifying their grid integration by allowing the TSO to schedule operations at specific time intervals, see Figure 10a; (ii) control the ramp rate of the generated power in case of sudden natural condition changes, see Figure 10b. A small-scale prototype was installed in Malta's Grand Harbor in May 2018. After one year, the testing campaign was completed. Results confirmed a consistently high thermal efficiency across a variety of meteorological conditions and operating regimes after hundreds of charging cycles. The prototype was removed and decommissioned, and nowadays the FLASC team is focused on developing a large-scale demonstrator in the open sea [181,182].

Conclusions
Future power scenarios include offshore wind energy as an important generation source. According to this framework, this paper discusses and reviews some aspects of offshore wind power plants for a massive integration into power systems. In the last decade, several characteristics such as offshore wind turbines, wind power plants, water depth and distance to shore have increased 230%, 700%, 170% and 110%, respectively. In the same way, electrical transmission has also evolved from HVAC to HVDC solutions. Moreover, HVDC technology currently offers three different possibilities: LCCs (based on thyristors), VSCs (based on IGBTs) and DRU (based on diodes). LCCs and VSCs have already been used, whereas DRU has not been implemented yet. The advantages and drawbacks of each technology have been extensively discussed in the paper. Different future advancements currently under development are also described: P2X conversion, the hub-and-spoke project as well as hydrogen and compressed air energy storage. The P2X conversion can enhance the power system's flexibility by converting the electricity surplus to other substances; however, its investment and maintenance costs should first be reduced to be economically viable. The hub-and-spoke project aims to facilitate the huge integration of offshore wind power plants in the North Sea; the total cost saving of this project, compared to a common offshore wind power plant, is estimated to be between 15 and 20 billion e. However, it is expected to take more than 10 years to become operational. Both hydrogen and compressed air energy storage systems appear as an alternative to conventional storage technologies due to the difficulty of using these traditional storage systems in the marine environment. Acknowledgments: This work was supported by 'Ministerio de Educación, Cultura y Deporte' of Spain (grant numbers FPU16/04282 and EST18/00738).

Conflicts of Interest:
The authors declare no conflict of interest.

Abbreviations
The following abbreviations are used in this manuscript: