**2. Looking into the future**

As discussed above, large offshore wind power plants (WPPs) are deemed suitable candidates to take up the responsibility of maintaining power system stability and security, potentially even participating in early state network restoration. Since the fast-growing capacity of the overall site and the individual turbines is pushing offshore WPPs further away from the shore and into deeper waters, as indicated in **Figure 3**, HVDC transmission is more suited to export the power to the onshore grid. Although more expensive, the fully controllable VSC interface allows HVDC to provide various dynamic grid support services that enhance system stability and resilience [5].

However, large offshore WPPs today consist of upto 100s of WTGs connected in a large inter-array network of upto 70 km of subsea cables, with long HVAC or HVDC transmission corridor that transports the bulk power onshore, requiring either special reactive power compensation or large converter substations both offshore and onshore to manage the power flow efficiently. This makes the offshore WPP an aggregated unit with a converter-dominated environment and a very rich resonance

<sup>6</sup> Like interconnectors, sites with trip-to-houseload operation and aggregated units such as wind and solar, supported by energy storage systems.

<sup>7</sup> Levelised cost of electricity.

<sup>8</sup> For example, refining and metallurgical industry, long-distance trucking and shipping.

#### **Figure 2.**

*Evolution of yearly average newly installed capacity of offshore wind turbines and farms; reproduced from [14]. Today the world's largest offshore wind farm is 1.2 GW Hornsea-1 (UK) and offshore wind turbines are already reaching ratings upto 15 MW, as of 2021.*

#### **Figure 3.**

*Average water depth and distance to shore of bottom-fixed offshore wind farms: the overall site capacity is indicated by bubble size; reproduced from [14].*

spectrum that must be first operated in a stable and robust manner before providing onshore grid services [17].

#### **2.1 Toward next-generation wind farms**

Traditionally, the first-generation grid-connected PEC-interfaced *grid-feeding* sources only supplied set-point based real and reactive power with basic survivability over a certain voltage and frequency range, while the second-generation devices now

**Figure 4.**

*Working philosophy of (a) grid feeding, (b) grid supporting and (c) grid forming units. Together (a) and (b) are referred to as grid following [18].*

are required to provide more support to the grid, especially in areas of high RES energy penetration, hence classified as *grid-supporting* units [6], shown in **Figure 4**. Today wind turbines and wind farms already contribute to system stability through provision of ancillary services such as fast power reduction as frequency response to create a spinning reserve margin for primary frequency regulation, steady-state and dynamic reactive power control for voltage support, and fault ride through characteristics that involve reactive current injection for improving voltage stability during faults. In countries of high RES penetration, more active participation of wind power in voltage and frequency stability is required. Thus, local vocal control in weak grid scenario and fast frequency response is being increasingly demanded by the new grid codes. Furthermore, latest requirements include power oscillation damping and synthetic inertia provision that mimics the exchange of kinetic energy from a synchronous rotating mass by injecting active power in proportion to calculated RoCoF [19].

However, while synthetic inertia is an important service to compensate for reduced inertia in the power system, it requires measurement-based activation unlike conventional inertia that is based on inherent physical characteristics of synchronous machines [20]. This makes it an insufficient replacement for inertia as true synchronous inertia-like response can be achieved only by *grid-forming* units [21]. Today wind turbines are *grid-following (GFL)* in that they rely on an external grid voltage<sup>9</sup> to which the control latches using a PLL10 for stable operation, effectively making the WTG behave as a current source injecting controlled power. Such wind turbines are currently exempt from network restoration services as not only can they not create their own voltage and lack controlled islanding capabilities but also connection in early stages of blackstart process can result in a recurrence of blackout due to the grid not being strong enough for large wind farms [17].

Now as the penetration of PEC increases, assumptions valid for stronger, traditional grids may no longer hold resulting in improper PLL behavior causing a negative impact of the controller power sharing and system stability [22]. This necessitates the next generation of converter-based units, called *grid-forming (GFM)* and shown in **Figure 4**, to be capable of proactively supporting the grid in all states, especially emergency and blackout without having to rely on services from synchronous generators. Thus, such GFM units must be able to take the lead in creating system voltage to *control instead of just supporting* its amplitude and frequency, prevent adverse control

<sup>9</sup> Either the main onshore grid or the offshore grid formed by the HVDC converter.

<sup>10</sup> Phase locked loop.

interactions, counter harmonics and unbalances and support system survival while contributing to short circuit power and system inertia—limited by the boundaries of energy storage capacity and available power rating [6]. These are key to ensure flexible, efficient and reliable operation of future decentralized converter-rich grids.

#### **2.2 Some potential advantages**

Contrary to conventional WTGs, a GFM wind turbine behaves as a voltage source, thus not only allowing outward energization without having to wait for completion of network reconstruction, but also potentially participating in sectionalizing strategy for defense against blackouts by ensuring continuity of power supply in a regional island or at the very least switching to trip-to-houseload operation that reduces restoration time compared with cold startup and facilitates bottom-up grid recovery [19].

Moreover, GFM-WTGs can potentially minimize dependence on the offshore auxiliary diesel generator that is associated with capital and operational costs11, not to mention the high emissions—especially when the diesel generator is operating at fullload during unscheduled outages that can last upto 4–6 months. Since a GFM-WTG can produce produce power to keep itself warm avoiding the risk to its health<sup>12</sup> as long as the wind blows, replacing the diesel genset with few of these can yield significant economic and reliability benefits over the project's lifetime [17]. The CO2 displacement can also contribute to reducing carbon-footprint taxes and ensure a smoother/ faster granting of permits in the future.

Additionally, GFM-WTGs will also come in handy in the future to supply charging stations offshore essential for maritime vessel electrification, thus displacing a significant amount of marine fuel with green electricity and ultimately playing a key role in achieving our climate goals. Thus, blackstart and islanding capabilities unlocked by GFM-WTGs are an essential feature of self-reliant WPPs that not only make them more actively participate in advanced voltage and frequency control but also enable them to take up the responsibility of ensuring stable and robust grid operation without relying on synchronous generation, thus accelerating net-zero transition.
