*2.1. Background of the Problem*

#### 2.1.1. Renewable Energy (RE) Planning and Wind Farm Construction

For both developing and developed economies, stable and adequate supplies of energy are mandatory. However, as world's fossil fuel-based energy resources are limited, renewable energy (RE) has emerged as a viable solution to replace conventional electric power generation [2,3]. In addition, utilising RE is also an 'environmentally sound' solution for a country to meet its sustainable development goals (SDGs) [4], this is also addressed by ESG (environmental, social, governance) targets that are established for business institutions [5].

Despite disagreements over the definitions of 'RE', 'green energy', 'sustainable energy', and 'clean energy', a classification of RE that is commonly accepted refers to the type of RE resource, including solar, wind, hydropower, geothermal, biomass, marine, and others [6–8]. This is due to the fact that the technological aspects to exploit these types of RE resource are usually totally different, thereby leading to salient 'watersheds' that can be told between them.

A stream of research adopting this view of classification is related to the 'portfolios', either in terms of selecting the optimal investment portfolio (by a company's decision makers) [9] or determining a country's optimal RE portfolio (by energy planners or operators within its long-term energy policy [10]. In research related to portfolios, uncertain or risky issues are also addressed [11,12]. However, in contrast to these operational management topics, more studies focus on the technologies to exploit various aspects of RE.

As an example, the technologies to exploit wind resources usually involve aspects of EE (electrical engineering), fluid dynamics/mechanics, and construction engineering (CE). Moreover, the first step in building a wind farm is typically to construct the foundations before erecting the wind turbines. For onshore or inland turbines, it is possible for the local weather conditions to be directly utilised as engineering parameters (e.g., steel structures, concrete materials, etc.). However, for offshore turbines, regardless of their working mode

(i.e., fixed or floating, depending on the water depth), underwater foundations are always required, which necessitates the well-proportioned HPC material used for grouting [13].

Constructing new wind farms is necessary for a country with unique conditions to exploit its wind resources. For offshore wind farms, in the Taiwan Strait, numerous projects have taken advantage of the very shallow waters of the 'Taiwan Bank' [14] and the high wind availability [15]; similar projects have been developed on the edges of the Mediterranean Sea [16,17]. For on-shore or in-land wind farms, projects have been initiated near the Sahara Desert, e.g., budget has been allocated for the construction of wind farms in Algeria [18].

The motivation of this study is to explore and better understand the parameters (or the parametric performance) of the HPC materials (the numerical values for which can be obtained experimentally) used for grouting the structural bases of offshore wind turbines. This may benefit the selection process before the real admixture of the HPC is determined and applied as grouting, and it may also generate a novel knowledge set to be used for anticipating the potential outcomes of experiments.

#### 2.1.2. The Special Weather/Sea Conditions in Taiwan and the Effects for Wind Turbines

As discussed in several studies, the soil of Taiwan is rich in sulphates and crystalline salts due to its geological features which result from frequent crustal deformation [19,20]. Osmosis of sulphate that occurs in concrete has caused serious problems for inland structures; the gypsum reaction is one such reaction, which can inflate the concrete and peel the surface off from the structure [1]. Therefore, anti-sulphate capability is typically a critical durability parameter in this context.

In addition, the high temperature and high humidity of Taiwan may accelerate these effects and decrease the life of structures. Moreover, the frequent earthquakes in the Circum-Pacific Belt and the typhoons that originated in the West Pacific region may also cause unexpected damage to structures. The damage is 'unexpected' because such events usually are inherently random, as is the scale of the event, e.g., the 921 Taiwan Earthquake of 1999, the Typhoon Morakot, etc. [21,22]. These types of natural disasters should be considered to be different from the damage caused by other types of unexpected events, such as fires [23].

These same conditions also apply to the Taiwan Strait, within which the Taiwan Bank is a very large but shallow continental shelf as a potential site for constructing offshore wind farms. Despite a high availability wind field above the sea all year, the land under water is susceptible to acid attacks (such as sulphates, chlorides, etc.), experiences the same extreme climate conditions discussed above (high temperature and high humidity), suffers from frequent natural disasters (earthquakes and typhoons) [24], and is scoured by the changing ocean currents throughout the year [25]. Therefore, quality HPC material is always required for wind turbines for its good workability during grouting, its beneficial mechanical properties during construction, and its long-term durability.

Thus, the parameters of an HPC sample (i.e., the experimental values from individual tests for all HPC samples which become 'variables' during data analysis) should be further studied and clearly understood. In addition, a suitable method is needed to determine the optimal HPC materials for construction based on the parametric data of the samples being tested during the experiments.

The parameters included in this study are organised and shown in Figure 2. It should be noted that, unlike in previous studies, electrical resistivity on surface (ERoS) is considered to be a durability property (parameter) of HPC samples in this research. This is believed to be a more reasonable consideration based on recent studies [26–28].

**Figure 2.** HPC Sample Parameters (Organised).

In Figure 2, the two fresh-property parameters are excluded from this study. These two tests are usually related to the workability of an HPC material during the engineering process, and their data formats are usually incompatible because the 'data sampling periods' of these two experiments are different than those in the other two categories. In contrast, the other eight variables have a quite large overlap in terms of their data sampling periods, so their data formats are compatible for the analysis (see Section 3.1).

In this study, the barrier between the hardened mechanical properties and durability properties is also removed intentionally. By doing so, not only can the pairs of parameters within a certain category be explored, but so can the pairs of parameters across two categories, so that relations between them and the intensities of the relations can be identified and evaluated. As will be shown later in Section 2.3, these relations include correlations, similarities, and predictive associations in terms of the ability to use one parameter to anticipate another (and how 'safely' it can be used).
