Sage Journals: Discover world-class research

Abstract

This study summarizes a microscale wind resource and site suitability assessments of Chatham and Waiheke Islands in New Zealand. It aims to provide information about the quality of the wind resource on these islands and to establish technical reasons for the failure of the wind power project on Chatham Island. The standard wind-industry software, Wind Atlas Analysis and Application Program (WAsP), was used to assess the wind resource, while WAsP Engineering and the WAsP Windfarm Assessment Tool (WAT) are used to assess the site suitability. The results from the observed wind climate showed mean wind speeds of 6.6 m/s and 5.5 m/s with power densities of 295 W/m² and 204 W/m² for the respective measuring stations on the islands. Using selected wind turbines on the islands revealed estimated mean wind speeds, power densities, and annual energy production (AEP) of 11.1 m/s, 1449 W/m², and 11.87 GWh/year and 7.1 m/s, 406 W/m², and 6.15 GWh/year for Waiheke and Chatham Islands, respectively. The islands possess Wind Power Class 6 and 7 wind resources suitable for utility-scale wind power applications. The total net AEP for the Chatham wind farm is 2.73 GWh/year, and the site suitability assessment (SSA) revealed an ill-suitability of siting wind turbines, given that the mean wind speed and the turbulence experienced exceed the standard IEC 61400 recommendations.

Keywords

Wind resource assessment Waiheke Island site suitability assessment WAsP WAT New Zealand

Introduction

The global issue of climate change has strengthened the focus on renewable energy installations worldwide. Among these, one of the fastest-growing energy sources for electrical power generation is wind energy with an installed capacity increase from 24 GW to 1021 GW over more than two decades from 2001 to 2023 (REN21, 2023; GWEC, 2023).

Several factors have made it necessary to further explore and harness wind energy including the historical trend and the future projections of growth; the increasing cost-competitiveness of large-scale wind projects; the implementation of stable policies to attain government-set renewable energy deployment targets; the growing need to address the global challenges of energy security, and environmental sustainability (REN21, 2023; NZ Ministry of Economic Development, 2011). This global growth is underpinned by widespread national exploration and installations.

Countries like New Zealand that are party to the Paris Agreement and the achievement of Article 2:1(a) keep “pursuing efforts to limit the temperature increase to 1.50 °C above pre-industrial levels” (United Nations, 2015). They also recognize the individual role they play in “significantly reducing the risks and the impacts of climate change” (United Nations, 2015). This approach, coupled with achieving Sustainable Development Goal 7, i.e., Affordable and Clean Energy by 2030, makes it imperative to “invest in solar, wind and thermal power” (UNDP, 2020). It is with this view in mind that further exploration of wind energy in New Zealand is scoped.

In 2023, renewable energy resources made up approximately 43% of the total primary energy supplied (TPES) in New Zealand. These energy resources are of essential importance in New Zealand because electricity generation is its largest user. The renewable share of net electricity generation was about 88% in 2023 (MBIE, 2024). Hydro-power generation still dominated (60%) electricity generation, but the trends indicate that this generation technology has matured and its gains will remain marginal (modest scale) (Kelly, 2011). Other sources of electricity generation included 18% from geothermal, 9% from gas, 7% from wind, 2% from Coal, 1% from solar and 2% from other sources (MBIE, 2024). This presents opportunities for growth for other renewable resources as the country works towards a 90% renewable generation mix by 2025 as envisaged by the Ministry for Business, Innovation and Employment (MBIE, 2017), or a 100% renewable future as modeled by (Mason et al., 2010). These opportunities have been aided by the national energy strategy, The New Zealand Energy Strategy, which focuses on the equal prioritization of diverse resource development, environmental responsibility, efficient energy use, energy security, and affordability (NZ Ministry of Economic Development, 2011). Wind energy has been, barring the dip in the generation mix between 2015 and 2017, on a growing trajectory since the first wind turbine installation. It has the potential to generate more than its current 7% capacity from the substantial wind resource that is available (Kelly, 2011). The substantial wind energy potential that New Zealand has was quantified following academic research in the 1970s and early 1980s (Kelly, 2011; Cherry, 1987; Dawber, 1988; Edwards and Dawber, 1980; Edwards et al., 1979; Edwards, 1990; Pyle, 2016; Dawber and Drinkwater, 1996). In the years that have followed, New Zealand has grown to have twenty-one (21) wind farms operational with an installed capacity of 1263 MW (2024) and about 750 MW of consented capacity currently under construction (New Zealand Wind Energy Association, 2025).

As the wind resource investigation and monitoring gained renewed momentum in the early 2000s, financed by private/independent companies in New Zealand, the link with academic research has suffered. There is a shortage of academic references for assessing optimal locations for siting wind turbines because “research on wind energy site location is still scarce” (Wei Zheng et al., 2019). This is referred to as “a lack of sufficient qualitative information on small or micro-scale wind energy projects that exist in New Zealand” (Reuther and Thull, 2011). Pyle's case study suggests that there is currently a poor connection between wind energy research and the industry; that “wind data for sites are commercially sensitive and a private good,” primarily because wind resource assessment is sponsored by and for private organizations (Pyle, 2016). Although industry experts assert that there is little need for academic research, given the depth of the understanding of New Zealand's wind regime, the lack of academic references detailing micro-scale site climatology beg to differ. A study on micro-scale classification of offshore wind energy in New Zealand has tried to address this problem to some extent (Wei Zheng et al., 2019).

Relatively few relevant published studies on wind resource assessment can be found in New Zealand which use the industry standard wind simulation software, WAsP to assess the wind resource on a microscale level. A study on wind power potential assessment for three locations in New Zealand used the RETScreen software for energy yield calculations focused on near coastal offshore wind resources reported Bluff Harbor having better wind energy potential then Wellington and Whangarei, using Global Wind Atlas Data available through the RETScreen software (Nasab et al., 2020). Another study on carbon and energy footprints of developing onshore wind farms in New Zealand reported that despite energy investments and GHG emissions during the production phase, the onshore wind plants offsets its emissions over its lifespan, making them suitable for the energy transition in New Zealand and elsewhere (Pincelli et al., 2025). In another study on an overview of the development and application of wind energy in New Zealand found that small-scale household wind turbines have notable power generation potential and economic potential in the long term (Zhang et al., 2023). A study on permitting the use of small wind turbines in rural New Zealand for best use of the wind resource reported recommendations that municipal councils could reduce barriers to small wind turbine installations by considering taller towers in the range of 15–20 m as acceptable for household based small wind turbine systems (White and Wakes, 2014). Another study on understanding New Zealand's wind resources as a route to 100% renewable electricity reported that with an installation of additional 4380 MW of wind farms together with current hydro productions, New Zealand could easily achieve the 100% target of generating electricity from renewable energy in the near future (Poletti and Staffell, 2021). Looking at the previous studies on wind energy in New Zealand, there is a gap in scientific literature in terms of site-specific and microscale wind resource assessments and site suitability assessment of potential wind farm sites in New Zealand.

This complements an opportunity to provide open, publicly available information about the quality of the wind resource at selected sites. Two locations, Chatham and Waiheke Islands, present unique conditions that warrant wind resource assessment. The locational remoteness of the former island, to New Zealand's North and South islands, makes separate power generation the only viable option on the island. Before 2010, the island's generating station consisted of five (5) diesel generators. However, in 2010, the Chatham wind farm was commissioned. This wind farm was mothballed in 2017, seemingly for financial (Underhill and NBR, 2017) and technical reasons. Reports of the wind farm shut down because of high wind speeds (Maslin and Herald, 2019) also present an opportunity to conduct a site suitability assessment (SSA). This is important because, regardless of how “rich” the wind resource is at a site, if the site conditions are not suitable for a wind power project, the said project will underperform and lead to financial losses.

Waiheke Island has seen an increased clamor for a wind farm, led by the Waiheke Initiative for Sustainable Energy's (WISE's) Waiheke Wind, seeking consideration because of “its location and the unique potential the island has for creating renewable energy” (Waiheke Initiative for Sustainable Energy, 2020). With Vector Limited gearing up to make Waiheke Island the world's first electric vehicle island (Vector, 2019), knowledge of its wind resource will be crucial in modeling interactions between wind energy generation and electric vehicle charging (Hindsberger et al., 2012).

Given the problems and opportunities described, there is a strong need to conduct a micro-scale wind resource assessment to estimate the resources on both the islands and to conduct a site suitability assessment for the sites with excellent wind resource and make open/available the findings. This study is also motivated by wind resource assessments done within the Oceania region using microscale and mesoscale numerical models like WindPRO, WAsP, and WRF (Dayal, 2015; Dayal, 2021; Sharma and Ahmed, 2016; Dayal et al., 2021a; Dayal et al., 2022; Dayal et al. 2021b; Dayal et al., 2024).

The objective of this study is to use ground-based observations from measurement masts in the de facto industry-standard microscale wind resource modeling tool, Wind Atlas Analysis and Application Program (WAsP), to predict and classify the available wind resources on Waiheke and Chatham Islands. And to use WAsP Engineering and WAsP Windfarm Assessment Tool (WAT) to carry out site suitability assessment of both the island locations. This study also aims to establish technical reasons for the failure of the wind power project on Chatham Island.

Section “Theoretical background” of this work provides the theoretical background, section “Methodology” describes the methodology, and the results are presented and discussed in section “Results and discussion,” while the conclusions are presented in section “Conclusions.”

Theoretical background

This section presents the underlying theory of the model implementation in WAsP, WAsP Engineering, and WAT used to compute the wind parameters relevant to the wind resource and site suitability assessments.

The WAsP model

To correct the impact of topography on measured wind data, WAsP implements these sub-models: stability, roughness change, shelter, and orographic models. The logarithmic law is used to model the wind profile over flat terrains at high wind speeds.

\begin{matrix} u (z) = \frac{u_{*}}{κ} l n (\frac{z}{z_{0}}) \end{matrix}

(1)

where

u (z)

is the wind speed at height z (a.g.l),

u_{*}

is the friction velocity,

κ

is the von Kármán constant (usually taken as 0.4), and

z_{0}

is the surface roughness of the homogeneous terrain. At moderate wind speeds, there are deviations to this law and thus a more general law, as shown below, is used, which considers atmospheric stability:

\begin{matrix} u (z) = \frac{u_{*}}{κ} [l n (\frac{z}{z_{0}}) - ψ (\frac{z}{L})] \end{matrix}

(2)

where

ψ

is an empirical function and L is the Monin-Obukhov length. These new parameter additions take into consideration the impact of surface heat flux.

The geostrophic drag law is used to model the geostrophic wind under neutrally stable conditions. It is thus expressed as:

\begin{matrix} G = \frac{u_{*}}{κ} \sqrt{{(l n (\frac{u_{*}}{f z_{0}}) - A)}^{2} + B^{2}} \end{matrix}

(3)

where G is the geostrophic wind, A and B are empirical constants (or functions), f is the Coriolis parameter, and other parameters remain as previously defined. Details are contained in Landberg et al. (1996) and Troen and Lundtang Petersen (1989).

Stability model

To model the stability fluctuations, WAsP treats the neutral state as a basic (average) state, while the non-neutral stability conditions are treated as perturbations about the mean. The effects of varying surface heat flux are captured by using models of the climatological average and root-mean-square of surface heat flux and are derived from the geostrophic drag law and the wind speed profile.

The procedure involves the first-order expansion of the previously listed equations, including the geostrophic drag law to derive an expression to evaluate the offset from the neutral state of surface heat flux. The collection of expressions thus generated is used to estimate the “contamination” degree in an input data set that is caused by the stability effects, as well as to reintroduce proper contamination values for varying heights and conditions. Details are contained in the European Wind Atlas project papers (Landberg et al., 1996; Troen and Lundtang Petersen, 1989).

Roughness change model

To model the effect of roughness, the roughness length ( $z_{0}$ ) of roughness elements is defined as a function of the height of the elements, the cross-sectional area ( $S$ ) facing the wind, and the average horizontal area ( $A_{H}$ ) over which these elements are distributed such that:

\begin{matrix} z_{0} = 0.5 \times \frac{h \times S}{A_{H}} \end{matrix}

(4)

Further details are contained in the European Wind Atlas project papers (Landberg et al., 1996; Troen and Lundtang Petersen, 1989).

Shelter model

The drag in the wind caused by obstacles, especially those close to the meteorological station, is modeled and corrected for by the following equations:

\begin{matrix} R_{1} = \frac{Δ u}{u} = 9.8 {(\frac{z_{a}}{h})}^{0.14} \frac{x}{h} (1 - P) η e x p (- 0.67 η^{1.5}) \end{matrix}

(5)

where:

\begin{matrix} η = \frac{z_{a}}{h} {(\frac{0.32}{l n (h / z_{0})} . \frac{x}{h})}^{- 0.47} \end{matrix}

(6)

and P is the porosity of the obstacle and lies between 0 and 1. It is the ratio of open to closed areas. h refers to the obstacle height,

z_{a}

the site (anemometer) height considered, x the downstream distance, and

R_{1}

the fractional reduction in wind speed for obstacles modeled as having infinite lateral dimensions (Troen and Lundtang Petersen, 1989).

Orographic model

The differences in terrain height around a meteorological station affect measured wind data at the said station. To correct the measured wind data, the orographic model is used. This is done in two steps. In the first step, the potential flow perturbations over the local terrain are modeled with:

\begin{matrix} \vec{u} = \nabla χ \end{matrix}

(7)

where

\vec{u} = (u, v, w)

is the three-dimensional velocity vector and

χ

is the potential. This shows the relationship between the velocity perturbations and the potential. The general solution to this is given by the equation below:

\begin{matrix} χ_{j} = K_{n j} J_{n} (c_{j}^{n} \frac{r}{R}) e x p (in ϕ) e x p (- c_{j}^{n} \frac{z}{R}) \end{matrix}

(8)

Here, $J_{n}$ is the nth order Bessel function, r is the radius, $ϕ$ is the azimuth, z is the height, $K_{n j}$ are arbitrary constants, and $c_{j}^{n}$ are the $i t h$ zero of $J_{n}$ . Further details can be found in the European Wind Atlas project paper (Troen and Lundtang Petersen, 1989) .

Statistics

The observed wind climate varies rapidly with time. This variation in measured wind speed data is captured meaningfully by averaging over a period T. The equations below capture the modeling of mean wind speed and power density by WAsP:

\begin{matrix} \bar{u} = \frac{1}{T} \int_{0}^{T} u (t) d t \end{matrix}

(9)

\begin{matrix} \bar{E} = \frac{1}{2} ρ \bar{u^{3}} \end{matrix}

(10)

The two-parameter Weibull distribution is used to collect, in a compact form, the wind data collected as shown in equation 11. It calculates the variation, skewness, and the frequency of the measured winds from which the available wind power density and wind speeds at which it is highest are computed (Chandel et al., 2014).

\begin{matrix} f (u) = \frac{k}{A} {(\frac{u}{A})}^{k - 1} e^{- {(\frac{u}{A})}^{k}} \end{matrix}

(11)

These sub-models and the statistical basis make up the analysis and application models of WAsP. The former is used for regional assessment and calculates the generalized wind climate (GWC). The latter is used for siting, where the actual yearly power production from specific wind turbines is calculated. To estimate the power generated at turbine site(s) and by wind farms, WAsP takes the power curve $P (u)$ of a wind turbine generator (WTG), and the probability density function $P r (u)$ (from the predicted wind climate computed with the application model) of wind speed at the hub height of the WTG and generates a power density curve. Integrating this curve produces the mean annual power produced (Landberg et al., 1996; Mortensen et al., 2019).

\begin{matrix} P = \int_{0}^{\infty} P r (u) P (u) d u \end{matrix}

(12)

For wind farms, like the one on Chatham Island, wake losses must be considered. With wind turbine hub height, rotor diameter, power curve, and the thrust coefficient curve, WAsP wind farm model computes the effective wind speed deficit $δ V_{01}$ of a downwind turbine using the equation:

\begin{matrix} δ V_{01} = U_{0} (1 - \sqrt{1 - C_{t}}) {(\frac{D_{0}}{D_{0} + 2 k X_{01}})}^{2} \frac{A_{o v e r l a p}}{A_{1}^{(R)}} \end{matrix}

(13)

where

D_{0}

is the upwind turbine rotor diameter,

U_{0}

is the unperturbed wind speed the upwind turbine experiences,

C_{t}

is the thrust coefficient,

X_{01}

the downwind horizontal distance between turbines and k is the wake decay constant (Mortensen et al., 2019).

WAsP engineering model

The LINCOM (LINearized COMputation) flow model (Astrup and Larsen, 1989) is implemented in WAsP Engineering and solves a set of linearized flow equations using a spectral solution method, i.e., in terms of Fourier transforms, in every mesh point of a rectangular grid. WAsP Engineering extends the capabilities of WAsP by focusing on determining/modeling load-generating parameters of the wind such as extreme winds, wind shears and profiles, and turbulence (DTU Wind Energy, 2016). Topographic data and the observed, generalized, or extreme wind climate are data inputs provided by WAsP for the computation.

The LINCOM flow, turbulence, and extreme wind models permit the generation of reports including the following: extreme wind report, detailed site report, turbulence for winds report, and extreme and mean climate site assessment report. These reports are used in WAT for detailed wind farm/site assessment.

Windfarm Assessment Tool (WAT)

The WAsP Windfarm Assessment Tool (WAT) functions to ease windfarm assessment based on IEC 61400-1 standard specification. With input data from WAsP and WAsP Engineering, a turbine classification system, site assessment protocol, and terrain complexity evaluation, the suitability for any site is determined (DTU Wind Energy, 2019).

Turbine classification

Classified according to wind speeds and turbulence that wind turbines are designed for, there are generally four wind speed classes, i.e., classes I–IV, and two turbulence classes, i.e., A (18%) and B (16%) (Rao, 2019) for each class. The turbine classification system considers the following wind speed (in m/s) parameters for each class: reference wind speed ( $U_{r e f}$ ), annual average wind speed ( $U_{a v e}$ ), 50-year return gust speed ( $1.4 U_{r e f}$ ), 1-year return gust speed ( $1.05 U_{r e f}$ ), at 1.225 kg/m³ air density (Gipe, 2006). Table 1 presents the wind turbine design classes.

Table 1.

Wind turbine design classes (Rao, 2019; Busby, 2012a, 2012b).

Class	Reference wind speed ( $U_{r e f}$ ) in m/s	Average wind speed ( $U_{a v e}$ ) in m/s
I	50	10
II	42.5	8.5
III	37.5	7.5
IV	30	6.0
S	Specified by designer

Turbulence is modeled with the IEC 61400-1 normal turbulence model (NTM) and based on editions (WAT has three—editions 2, 3, and 4). Characteristic/representative turbulence intensities are defined as functions of the mean and standard deviation of random ten-minute samples/measurements (DTU Wind Energy, 2019). Depending on the edition, the turbulence classification extends beyond the two classes stated by (Rao, 2019), to as many as five (A+, A, B, C, S), as contained in IEC 61400-1 edition 4.

IEC 61400 site assessment criteria

To ensure the safe deployment of wind turbines, a list of criteria is defined, that should be met, in the IEC 61400-1 standard Wind turbines, Part 1: Design requirements. WAT's turbine classification system compares the extreme and average wind speeds, and the mean, effective, and extreme turbulence intensities to recommend an IEC turbine class most suitable for use at turbine sites. The flow conditions present a summary of the major criteria, and conditions, to be met including the 50-year extreme wind speed ( $U_{50, r e f}$ ), the average wind speed ( $U_{a v e}$ ), the flow inclination ( $φ$ ) relative to the centerline of the nacelle, the allowable wind shear exponent ( $α$ ), the mean $I_{15} (u)$ , effective $I_{e f f} (u)$ , maximum $I_{m a x} (u)$ , and extreme $I_{e x t} (u)$ , turbulence intensities, the allowable wind speed distribution $p (u)$ , the allowable Weibull-k parameter, and an assessment of terrain/topographic complexity. A summary of the criteria is shown in Table 2.

Table 2.

Summary of site assessment criteria (DTU Wind Energy, 2019).

	Criteria	Condition (for Edition 4)
1	Extreme wind speed ( $U_{r e f})$	$ρ_{s i t e} U_{50, s i t e}^{2} \leq ρ_{d e s i g n} U_{r e f}^{2}$
2	Flow inclination ( $φ$ )	$φ \leq 8 \circ$
3	Allowable wind shear ( $α$ )	$0.05 \leq α \leq 0.25$
4	Allowable (normal) turbulence intensity ( $I (u)$ )	$(C_{c t} I_{e f f, 90 %} (u) \leq N T M (u))$ $f o r (0.6 U_{r a t e d} \leq u \leq U_{c u t - o u t}) o r (U_{a v e} \leq u \leq 2 U_{a v e})$
5	Allowable wind speed distribution ( $p (u)$ )	$p (u)_{s i t e} \leq p (u)_{d e s i g n}$ $f o r (0.6 U_{r a t e d} \leq u \leq 1.6 U_{r a t e d}) o r U_{a v e} \leq u \leq 2 U_{a v e}$
6	Weibull shape parameter limitation	$(6.5 x - 4.5) \leq k \leq (8.0 - 6.0 x)$
6	Weibull shape parameter limitation	$k \geq 1.4$
7	Air density	$ρ_{a i r} \leq 1.225 k g / m^{3}$
8	Maximum wake turbulence	$C_{c t} I_{m a x} (u) \leq E T M (u)$ $f o r (0.6 U_{r a t e d} \leq u \leq 1.6 U_{r a t e d})$
9	Extreme ambient turbulence	$I_{e x t r e m e} (u) \leq E T M (u)$ $f o r (0.6 U_{r a t e d} \leq u \leq 1.6 U_{r a t e d})$

where $x = \frac{U_{a v e, s i t e}}{U_{a v e . d e s i g n}}$ , and $C_{c t}$ is a correction parameter used in edition 4 to correct for complex terrain. A site is considered good for siting wind turbines if all the criteria are met. It is important to note that some criteria will not be met, for which modifications either from the turbine manufacturer (concerning speed distribution and effective turbulence plots) or from wind farm operation (WAsP DTU Wind Energy, 2015a) can be undertaken to make seemingly unsuitable turbine sites fit for siting (WAsP DTU Wind Energy, 2015b). This paper does not broach those topics.

Methodology

In this section, the study locations, meteorological data used and the modeling approach with WAsP, WAsP Engineering and the Windfarm Assessment Tool (WAT) are described.

Study locations and meteorological data

Figure 1 shows the study locations of Chatham and Waiheke Islands in New Zealand with earmarks on the location of the meteorological stations on both islands. The meteorological data for Chatham Island consists of nineteen (19) years of hourly averaged wind speed and direction measured at an elevation of 12 m above ground level (a.g.l), collected over the period spanning from October 1999 to November 2018. This data was supplied by the National Institute of Water and Atmospheric Research (NIWA) for Tuuta meteorological station whose coordinates are (−43.81687, −176.475). Information about the station listing is shown in Table 3.

Figure 1.

Chatham (up) and Waiheke (down) Islands (Source: Google Maps).

Table 3.

Station listing for Chatham Island.

Agent No.	Network No.	Start date	End date	Percent complete	Name	Latitude	Longitude
Agent No.	Network No.	Start date	End date	Percent complete	Name	(decimal degrees)
17840	K98616	12-Oct-1999	07-Nov-2018	100	Chatham Island EWS	−43.8168	−176.475

Note: EWS: electronic weather station.

For Waiheke Island, MetraWeather (MetraWeather, 2025) provided eleven (11) years of hourly averaged measured wind data collected over the period spanning from December 2008 to December 2019. The meteorological station (−36.85, 175.117) at Passage Rock measured wind data at a height of 8 m above sea level (a.s.l).

Software setup and modeling

WAsP

Modeling with WAsP typically involves wind data analysis, defining topographical concepts for the roughness, shelter, and orography models, the estimation of power produced at a turbine site, as well as the power production from a wind farm. Prior to setting up a WAsP project, the application provides some utility tools to prepare meteorological and topographic input data into the right formats. Wind data analysis was done by the WAsP Climate Analyst (WACA 3.1). It performed analyses on the time series of meteorological wind data and produced a summary of extreme and mean wind climates. A WACA project was set up for the met stations on Chatham and Waiheke Islands from where data was collected. Sites and instruments were set up, and data were imported, selected, and mapped onto instrument heights, for the calculation of an observed mean wind climate (OMWC) which describes the overall mean level of wind resource available at a site with temporal variation, that was exported to WAsP. An observed extreme wind climate (OEWC) describing the magnitude of the strongest winds was also exported for use in WAsP Engineering later.

The WAsP Map Editor 12.3 was used which allows the creation, editing, and inspection of topographical maps in WAsP. These maps contain terrain information like elevation and roughness, and this tool allows the terrain roughness and elevation data to be imported from web databases. For this article, the database used was the Global Wind Atlas (GWA) Map Warehouse. After separate importation of roughness and terrain data, they are added together to form a single digitized map containing elevation and roughness for both islands as shown in Figure 2.

Figure 2.

Digitized maps of Chatham (up) and Waiheke (down) Islands showing combined elevation and roughness maps.

With the OMWC and the digitized maps, a WAsP project was set up for both islands. Inside these projects, the obstacles were defined close to meteorological stations at Passage Rock (Waiheke Island) and Tuuta Airport (Chatham Island), and a generalized wind climate (GWC) was calculated.

The WAsP model uses the Wind Atlas Methodology, which uses the measured wind data to generate a site-independent generalized wind climate (wind atlas). The generated wind atlas represents the wind climate over ideal, featureless terrain and is then used with combined roughness and topographic maps of a specific site to calculate the actual, site-specific wind climate. This is done by analyzing the measured wind data, then transforming them into a generalized wind atlas and then applying inverse calculations with local terrain and roughness data to predict wind conditions at any location under study.

To calculate the annual energy production over the entire islands (as in a resource grid), or at given site locations, appropriate wind turbines must be selected. For Chatham Island, initial turbine selection was guided by the existing 225 kW Vergnet GEV MP-C 30 turbines (McWhirter, 2010) mothballed on the island. Given the information provided by Vergnet, the GEV MP-C 30 m twin-rotor turbines were installed (Vergnet Wind Turbines, 2020). These WTGs generally have power ratings ranging from 200 kW to 275 kW (Vergnet Wind Turbines, 2014). These sources form the bases from which the power and thrust coefficient data used in modeling were created. Figure 3 (left) shows the plot of these curves.

Figure 3.

Thrust coefficient and power curves for Vergnet GEV MP 30 (left)^a turbine used on Chatham Island and Siemens SWT-2.3-82 (right)^b turbines selected for Waiheke Island.

For Waiheke Island, there has been no wind farm or turbine(s) sited, hence initial selection will be guided by the trend in existing wind farms in New Zealand. Of the six hundred and sixty-four (664) turbines installed in New Zealand, the 2.3 MW Siemens SWT-2.3-82 accounts for majority of the wind turbines. This type was selected for the resource assessment on Waiheke Island. Figure 3 (right) shows the WTG curves selected for Waiheke Island.

Note that site suitability assessment will provide better insights into the wind loading factors that affect the wind turbines selected. From the assessment, final turbine selections/recommendations are (can be) made. With the turbine selected and added to a turbine site or windfarm, mean speed at hub height, the total gross and net annual energy production (AEP), power density, and others are calculated for the resource grid using WAsP 12.1. The resource grid presents a spatial variation of the wind resource, as well as the ruggedness index (RIX), which indicates the degree to which terrain complexity (orography), outside WAsP operational envelope, impacts the accurate prediction of the wind resource (Mortensen et al., 2019).

WAsP Engineering

A WAsP Engineering project was set up for Chatham and Waiheke Islands, each by first inputting the project domain of elevation and roughness maps. This map was obtained from WAsP Map Editor. Then, the observed extreme wind climate (OEWC) and/or observed, generalized, or geostrophic winds are input into WAsP Engineering. Windfarms are then inserted from the site locations with class 6 or 7 wind resource. For Chatham Island, the windfarm layout remains as it is, while for Waiheke Island, as much as is possible within areas with excellent wind resource, the rectangular array recommendation by Busby (2012a, 2012b) is adopted. Modifications are made, as deemed appropriate, to create a best-fit layout. For all winds (observed, generalized, and geostrophic), WAsP Engineering solves and reports for a list of parameters including wind speeds, roughness, friction velocities, “u,” “v,” and “w” component velocities and turbulent intensities, roughness, and flow inclination, amongst others. For every turbine site in each windfarm, at every defined height (usually the mid-lower, exact, and mid-upper hub heights), and given the GWC of both islands and selected wind turbines, a site assessment report is generated.

Windfarm Assessment Tool (WAT)

The site assessment report generated for Chatham and Waiheke Islands is imported into WAT, and a new project was set up for both islands, separately. In this WAT projects, the flow conditions at each turbine site are presented with green-filled circles indicating that a flow condition at a site is “OK” and a red-filled circle indicating a “problem”. Because the majority of the calculations for site suitability are done in WAsP and WAsP Engineering, WAT presents an easier-to-manage overview with the option to modify certain parameters, like sector management, as well as modeling of the mean and the standard deviation of the turbulence intensities (DTU Wind Energy, 2019).

For Chatham Island, the site suitability report for selected excellent sites, including the Chatham windfarm, is presented. For Waiheke Island, the report covered sites with excellent wind resources.

Figure 4 shows the pictorial representation of the processes this paper takes to estimate the wind resource on Chatham and Waiheke Islands and to conduct the site suitability analysis (SSA) on sites with excellent utility-scale wind resource.

Figure 4.

A pictorial representation of WAsP analysis (in bounded box) (Troen and Lundtang Petersen, 1989) and the modification to include WAsP engineering and WAT modeling.

Results and discussion

In this section, the simulation results from WAsP, WAsP Engineering, and WAT are presented. For WAsP, the temporal variation of the observed mean wind climate (OMWC) for both islands are presented. The spatial variation of the wind resource, as contained in the resource grids of the mean wind speed, annual energy production (AEP), and power density, are presented for both the islands. For Chatham Island, which has an existing but mothballed wind farm, the AEP for the said wind farm is also highlighted and discussed.

Finally, the results of the site suitability assessment (SSA) for regions/sites on both these islands with excellent wind resources (isolated and fitted with wind farms) are presented and discussed. These results are obtained from a simulation with WAsP Engineering and WAT.

Temporal variation

In this subsection, a summary of the time-independent observed mean wind climates (OMWC), diurnal and annual variation of measured wind for Chatham and Waiheke Islands is presented.

Observed mean wind climate

The OMWC for Chatham Island is shown in Figure 5 and summarizes the sector-wise and overall observations for Chatham Island. The most frequent mean wind speeds are experienced in sectors 1 (11.1%), 8 (13.6%), 9 (15.1%), and 12 (12.6%). The corresponding sector angles and wind speeds are 6.81 m/s at 0°, 7.36 m/s at 210°, 7.21 m/s at 240°, and 6.66 m/s at 330°. The prevailing winds are south-westerly and north-westerly. The all-sector omni-directional mean wind speed is 6.59 m/s and the power density for this speed at 1.225 kg/m³ density is 295 W/m². The fitted all-sector Weibull-A parameter is 7.5 m/s, while the Weibull-k parameter is 2.32. The wind resource parameters at the Chatham Island measurement station are in the Wind Power Class 3 and above category, suitable for utility-scale wind power projects.

Figure 5.

Wind rose (up) and histogram (down) of observed wind data measured on Chatham Island.

Similarly, for Waiheke Island, the OMWC is shown in Figure 6. It summarizes the sector-wise and overall observations for Waiheke Island. The most frequent mean wind speeds are experienced in sectors 8 (14.6%), 9 (16.9%), and 4 (10.1%). The corresponding sector angles and wind speeds are 5.66 m/s at 210°, 6.80 m/s at 240°, and 5.75 m/s at 90°. The prevailing winds are south-westerly. The all-sector omni-directional mean wind speed is 5.48 m/s and the power density, for this speed at 1.225 kg/m³ density is 204 W/m². The fitted all-sector Weibull-A parameter is 6.2 m/s, while the Weibull-k parameter is 1.94. The wind resource parameters at Waiheke Island measurement station are in Wind Power Class 3 and above category suitable for utility-scale wind power projects.

Figure 6.

Wind rose (up) and histogram (down) of observed wind data measured on Waiheke Island.

Diurnal variation

From the omni-directional wind statistics, the hourly variation of observations for each month is analyzed. Figure 7 shows a graphical representation of the diurnal variation of measured wind speed on Chatham Island at 12 m a.g.l., which ranges between 5.9 m/s and 7.7 m/s. The pattern, for each month, is a distinct “Gaussian” with peaks between 12:00 and 14:00 h. The diurnal variation of wind speed at Chatham Island measurement station is greater than the cut-in wind speed of 3.5 m/s of the Vergnet GEV MP 30 wind turbine and thus turbines sited will always be in production throughout the day of every month and year. The pattern of the diurnal variation of wind speed is similar to potential wind farm locations within the Oceania region (Dayal et al., 2021a; Dayal et al., 2021b).

Figure 7.

Diurnal variation of observations on Chatham Island.

The hourly variation of observations for each month on Waiheke Islands at a height of 8 m a.s.l. is shown in Figure 8. The mean wind speed ranges between 4.7 m/s and 6.6 m/s. The pattern, for each month, is distinctly “sinusoidal” with peaks between 02:00 and 06:00 h.

Figure 8.

Diurnal variation of observations on Waiheke Island.

The diurnal variation of wind speed at the Waiheke Island measurement station is also greater than the cut-in wind speed of 3.5 m/s of the Siemens SWT-2.3-82 wind turbine, and thus turbines sited will always be in production throughout the day of every month and year. The pattern of the peak diurnal variation of wind speed is different from Chatham Island due to the location of the measurement station, which is very close to the sea.

Annual variation

The annual variation on Chatham Island is jagged with mean wind speed hovering (generally) around a mean of 6.6 m/s, as shown in Figure 9. The annual mean wind speed ranges from 5.7 m/s to 7.4 m/s. The annual variation of wind speed at the Chatham Island measurement station is greater than the cut-in wind speed of 3.5 m/s of the Vergnet GEV MP 30 wind turbine, and thus turbines sited will always be in production throughout every month of every year except in 2016 for two months.

Figure 9.

Annual variation of observations on Chatham Island.

Waiheke Island's variation is similarly jagged with mean wind speed hovering around a mean of 5.3 m/s as shown in Figure 10. The mean wind speeds range from 4.9 m/s to 6.8 m/s. Seasonal variation over each year shows that spring and summer experience higher mean wind speeds. The annual variation of wind speed at the Waiheke Island measurement station is greater than the cut-in wind speed of 3.5 m/s of the Siemens SWT-2.3-82 wind turbine, and thus turbines sited will always be in production throughout every month of every year.

Figure 10.

Annual variation of observations on Waiheke Island.

Spatial variation

The spatial variation of the wind resource on Chatham and Waiheke Islands is presented in 2-D contour maps at horizontal resolutions of 125 m × 125 m and 60 m × 60 m, respectively. The maps depict changes in mean wind speed, annual energy production (AEP), and the power density as calculated in the resource grids at selected wind turbine hub heights of 55 m and 80 m, respectively.

Mean wind speed

The mean wind speed on Chatham Island ranges from 4.8 m/s to 14.80 m/s, as shown in Figure 11. A significant fraction of the landmass has wind speeds above 8 m/s. Over the entire island, the mean wind speed is 11.08 m/s. The minimum mean wind speed is 4.88 m/s and occurs at (536771, 5116985). The maximum mean wind speed is 14.80 m/s and occurs at (529646, 5115110). The spatial variation of minimum, maximum, and mean wind speeds across Chatham Island is greater than the cut-in wind speed of 3.5 m/s of the Vergnet GEV MP 30 and other wind turbines, and thus turbines sited anywhere across Chatham Island will always be in production.

Figure 11.

Contour map of mean wind speed variation over Chatham Island.

The mean wind speed on Waiheke Island ranges from 5.3 m/s to 8.99 m/s, as shown in Figure 12. Wind speed over 8 m/s is sparsely distributed across the island. Over the entire island, the mean wind speed is 7.14 m/s. The minimum mean wind speed is 5.30 m/s and occurs at (332867, 5923972). The maximum mean wind speed is 8.99 m/s and occurs at (332167, 5924322). The spatial variation of minimum, maximum, and mean wind speed across Waiheke Island is greater than the cut-in wind speed of 3.5 m/s of the Siemens SWT-2.3-82 and other wind turbines, and thus turbines sited anywhere across Waiheke Island will always be in production.

Figure 12.

Contour map of mean wind speed variation over Waiheke Island.

Power density

Figure 13 shows the spatial variation of the wind power density on Chatham Island. The power density ranges from a minimum of 189 W/m² at (536646, 5116985) to 3760 W/m² at (540396, 5118485). The mean power density is 1449 W/m². The spatial variation of minimum, maximum, and mean power density across Chatham Island is favorable for the Vergnet GEV MP 30 wind turbine, and the mean power density is greater than 300 W/m², which indicates that the wind resource can be classified as Wind Power Class 3 and above across Chatham Island.

Figure 13.

Contour map showing the power density variation over Chatham Island.

Similarly, the power density on Waiheke Island ranges from a minimum of 167 W/m² at (332867, 5923922) to 766 W/m² at (332167, 5924322). The mean power density is 406 W/m². Figure 14 shows the spatial variation of power density on Waiheke Island.

Figure 14.

Contour map showing the power density variation over Waiheke Island.

The spatial variation of minimum, maximum, and mean power density across Waiheke Island is favorable for the Siemens SWT-2.3-82 wind turbine, and the mean power density is greater than 300 W/m², which indicates that the wind resource can be classified as Wind Power Class 3 and above across Waiheke Island.

Annual energy production

Figure 15 presents the spatial variation of the AEP over Chatham Island. The minimum AEP computed at 55 m hub height with Vergnet GEV MP-C 30 m rotor WTG is 3.010 GWh/year, occurring at (536646, 5116985), while the maximum is 13.873 GWh/year at (531521, 5114360). The mean AEP over the entire resource grid is 11.873 GWh/year. The spatial variation of minimum, maximum, and AEP across Chatham Island is favorable for the Vergnet GEV MP-C 30 wind turbine, and the AEP is greater than 0.500 GWh/year, which indicates that the wind resource can be classified as Wind Power Class 3 and above across Chatham Island.

Figure 15.

Contour map showing the AEP variation over Chatham Island.

Figure 16 shows the spatial variation of the AEP at Waiheke Island. The mean AEP is 6.149 GWh/year calculated with 80 m hub height Siemens SWT-2.3-82 WTG. The minimum AEP is 2.931 GWh/year occurring at (332867, 5923972), while the maximum is 9.219 GWh/year at (332167, 5924322). The spatial variation of minimum, maximum, and mean AEP across Waiheke Island is favorable for the Siemens SWT-2.3-82 wind turbine, and the mean AEP is greater than 0.500 GWh/year, which indicates that the wind resource can be classified as Wind Power Class 3 and above across Waiheke Island.

Figure 16.

Contour map showing the AEP variation over Waiheke Island.

Wind farm production

Chatham wind farm has two turbine sites located at positions shown in Table 4. Both these turbines combine to produce 2.779 GWh/year of total gross AEP and 2.727 GWh/year of total net AEP with a proportional wake loss of 1.85% as shown in Table 5. The mean wind speed at Chatham A turbine site is 10.26 m/s, with a power density of 1147 W/m² and a net AEP of 1.382 GWh/year. The wake loss at this site is 0.92%. Similarly, for the Chatham B turbine site, the mean wind speed at this turbine site is 10.17 m/s, with a power density of 1113 W/m², net AEP 1.346 GWh/year, and wake loss of 2.8%.

Table 4.

Chatham windfarm turbine sites.

Site description	X-location [m]	Y-location [m]
Chatham A	549356.9	5123892
Chatham B	549410.7	5123797

Table 5.

Chatham wind farm production.

Variable	Total	Mean	Min	Max
Total gross AEP [GWh]	2.779	1.389	1.384	1.395
Total net AEP [GWh]	2.727	1.364	1.346	1.382
Proportional wake loss [%]	1.85	-	0.92	2.8
Mean speed [m/s]	-	10.21	10.17	10.26
Mean speed (wake-reduced) [m/s]	-	10.11	10.02	10.21
Air density [kg/m³]	-	1.212	1.212	1.212
Power density [W/m²]	-	1130	1113	1147
RIX [%]	-	-	1.8	2

Following the mini-report (McWhirter, 2010), the annual power required in Chatham's power generation system is 2.2 GWh/year. Based purely on resource potential, the 2.727 GWh/year net AEP of the Chatham wind farm can match the power requirements on the island. However, a study on modeling of the wind flow over Chatham Island reported estimation of power output from a single turbine was 125 kW, 55% of the maximum power rating of 225 kW. This study also highlights that the actual power output is much lower than the statistical estimation. Operational reasons were attributed to the low turbine performance (Tonkin et al., 2017). However, with average wind speeds of 10.26 m/s and 10.17 m/s, which are higher than the highest IEC wind turbine class, i.e., class I, as well as reports of turbine shutdown when speeds exceed 20 knots, i.e., 10.28 m/s, as reported by (Maslin and Herald, 2019), the suitability of the site must also be considered. This justifies the need for the site suitability assessment (SSA) (Figure 17).

Figure 17.

Close-up view of Chatham windfarm location with two turbine sites (+).

Resource classification

Different wind classification systems, based on wind speed and power density, exist. The wind potentials on Chatham and Waiheke Islands are classified based on the seven-class system detailed in (Busby, 2012a, 2012b). Figure 18 and Figure 19 show the mean wind speed and power density across Chatham Island equal to and greater than 8.7 m/s and 800 W/m², respectively, at 55 m turbine hub height. Following resource classification detailed in (Busby, 2012a, 2012b), a large portion of the island has a Wind Power Class 7 wind resource. This is considered as excellent utility-scale potential.

Figure 18.

Class 7 mean wind speed distribution across Chatham Island. The red circle highlights the Chatham windfarm. White space shows sites with wind speeds lower than 8.7 m/s.

Figure 19.

Class 7 power density distribution across Chatham Island. White space shows sites with wind power density lower than 800 W/m².

The Chatham wind farm with turbine sites located as shown in Figure 17 was sited where the mean speed for a 55 m hub-height WTG was 10.21 m/s. With the same 55 m Vergnet GEV MP WTG, the power density (see Figure 19) across the island ranges from 800 W/m² to 3800 W/m².

Similarly, the Waiheke Island spatial charts (Figure 20 and Figure 21) below show a sparser distribution of sites with excellent potential, i.e., Class 4 (power density ranging between 400 and 500 W/m² and wind speed between 7.0 and 7.5 m/s) and higher. The distribution shows where the Class 6 (power density ranging between 600 and 800 W/m² and wind speed between 8.0 and 8.8 m/s) wind resource lies.

Figure 20.

Sites with Wind Power Class 6 power density on Waiheke Island.

Figure 21.

Sites with Wind Power Class 6 mean wind speeds on Waiheke Island.

Turbine sites within these distributions were selected, at the modeler's discretion, for site suitability assessment (SSA). This discretion is guided mostly by the recommended minimum side-by-side spacing of five rotor diameters between turbine sites in a wind farm (Busby, 2012a, 2012b). The turbine arrangements are mostly in irregular arrays or follow a terrain feature with positions hand-drafted, as was done in (Mortensen et al., 2012).

Site suitability assessment

The site suitability assessment (SSA) results for both islands are presented below. For Chatham Island, SSA focused on the Chatham wind farm as well as potential sites identified with excellent wind resources across the island, as shown in Figure 22. For Waiheke Island, in the absence of any existing wind project, the SSA focused on sites with excellent potential.

Figure 22.

Potential wind farms with turbine sites distributed over different sites with class 7 wind resources on Chatham Island. The Vergnet GEV MP-C 30 WTG with a hub height of 55 m was used.

For the Chatham windfarm, in particular, Table 6 outlines the flow conditions calculated in WAsP Engineering and assessed with WAT. The 50-year extreme wind ( $V_{50}$ ) for Chatham A and Chatham B turbine sites are 35.72 m/s and 35.56 m/s, respectively. Both these values are Class III because they are below the 37.5 m/s IEC 61400 extreme wind speed of the said class. Class III WTG are designed to experience a mean wind speed of 7.5 m/s; however, the average wind speeds ( $V_{a v e}$ ) at hub height, i.e., 10.26 m/s and 10.17 m/s, are both higher than the standard specification for class III, II, and for the highest class I (10 m/s) as well.

Table 6.

Site suitability assessment results of Chatham windfarm.

IEC 61400 site assessment criteria Flow conditions	Turbine site(s)
IEC 61400 site assessment criteria Flow conditions	Chatham_A	Chatham_B
IEC class	IA+ (ED4)	IA+ (ED4)
$V_{50}$ [m/s]	35.72	35.56
$V_{a v e}$ [m/s]	10.26	10.17
Shear exponent ( $α$ )	0.152	0.168
Flow inclination angle [deg]	0.82	0.36
Mean $I_{15}$	17.17%	17.95%
$I_{e f f}$	problem	problem
$I_{e x t r e m e, a m b i e n t}$	problem	problem
$I_{m a x}$	OK	problem
Speed pdf $p (u)$	problem	problem
Weibull k	problem	problem
Topography	OK	OK
$C_{c t}$	1	1

The wind shear exponents of 0.152 and 0.168 and flow inclination angles of 0.82° and 0.36° are both within the stipulated ranges. Assessment of the complexity of the surrounding terrain found topography to be “OK,” explaining why the turbulence structure correction factor ( $C_{c t}$ ) is unity (1).

Apart from the average wind speed ( $V_{a v e}$ ), other site conditions are more severe than the site assessment criteria. The “problem” entries in Table 6 indicate a failure to meet the specified criteria. The effective turbulence intensity ( $I_{e f f}$ ), which should be lower than the turbulence intensity as modeled by NTM over the specified range of 0.6–1.6 of the rated speed, are not met. The same goes for the extreme ambient turbulence intensity, speed distribution, Weibull-k parameter, and maximum turbulence intensity for Chatham B.

Figure 23 shows the turbulence intensity (top) and wind speed probability distribution (bottom) curves for the Chatham A turbine site. As can be seen in the top figure, the all-sector effective turbulence intensity curve lies above the NTM A+ turbulence category. The A+ category is the highest (18%) measure that Class I WTG is expected to experience. This alludes to the WTGs at the Chatham wind farm experiencing turbulence higher than they are designed for.

Figure 23.

Top: Site and IEC turbulence intensity curves for Chatham A. Bottom: Site and IEC wind speed probability distribution curves.

Figure 23 (bottom) shows the comparison between the IEC and site probability distribution ( $p d f$ ) curves. The $p d f$ for the site, i.e., Chatham A, (blue curve), is sometimes above the IEC $p d f$ (red curve) in the specified $p d f$ range. This is problematic because the WTG is experiencing speeds higher than it has been designed for. (WAsP DTU Wind Energy, 2015b) states that, although this is an SSA flag, it can be remedied by WTG manufacturer modifications, and not by any other modification that wind resource analysts can conduct.

Similar information exists for Chatham B. From the assessment, no one site was found to meet all the SSA criteria for the Vergnet WTG used in the modeling. Considering the excellent nature of the wind resource on Chatham Island, a special class of WTG with a lower hub height than Vergnet WTG, designed to address the challenges with turbulence and speed distribution is recommended as a replacement. Another option is to relocate the wind farm to a lower mean wind speed area further inland.

On the other hand, in the absence of an existing wind project on Waiheke Island, wind farms have been sited as shown in Figure 24. The flow conditions show 14 windfarms with differing numbers of turbines being sited in regions with Wind Power Class 6 wind resource. Of these sites, only three show the most potential. These wind farm sites meet all SSA criteria except the probability distribution function $p d f$ criteria, which can be handled by WTG manufacturer modifications.

Figure 24.

Potential wind farms with turbine sites distributed over different sites with class 6 wind resource on Waiheke Island. The Siemens SWT-2.3-82 WTG with a hub height of 80 m is used.

For most of the turbine sites, a class IIC turbine classification is sufficient because extreme wind speeds hover at or below 42.5 m/s, average wind speeds at or below 8.5 m/s, and turbulence intensities below 14% which are well within the IEC criteria.

Limitations

The resolution of the resource grids for Chatham and Waiheke Islands was limited to 125 m and 60 m. At resolutions lower than these, WAsP failed to solve, even with an increase in PC processing power. This limitation is probably a result of some built-in calculation constraints on student licenses or the computation power of the computer used. This same limitation was experienced with WAsP Engineering's domain structure resolution. A finer resolution size of 10 m, the initial project selection, would have been preferred.

These islands had only one meteorological station observing the wind. This made cross-prediction and cross-correlation analysis for the verification of the resource prediction accuracy impossible.

With the Chatham wind farm mothballed, there was no access to actual production data, or information on operational configurations, as was the case with the study on modeling wind flow over Chatham Island (Tonkin et al., 2017). Therefore, comparison of actual production data with simulation results could not be performed.

The simulation results may improve if wind data measurements at higher heights and closer to the hub-height of 55 m and 80 m would be used. Wind measurements at higher heights will reduce the effect of ground friction on the wind speed in comparison with wind speeds at lower heights of 10 m. Higher resolution elevation and roughness data than those used in this study via vector maps may also improve the simulation results.

Conclusions

A micro-scale wind resource and site suitability assessment (SSA) was conducted on Chatham and Waiheke Islands using three modeling tools: WAsP, WAsP Engineering, and Windfarm Assessment Tool (WAT). Observed wind data collected at Tuuta Airport (Chatham) and Passage Rock (Waiheke) over a nineteen (19) and eleven (11) year period, respectively, were used for the analysis.

The observed mean wind climate (OMWC) on Chatham Island depicts an all-sector omni-directional mean wind speed of 6.59 m/s and power density of 295 W/m², with dominant south-easterly and north-easterly winds. The diurnal variation ranges between 5.9 m/s and 7.7 m/s. Seasonal variations depict spring and summer as the best performing seasons with peaks above 7.5 m/s at midday. The mean wind speed for Waiheke Island is 5.48 m/s with a power density of 204 W/m². The diurnal variation ranges between 4.7 m/s and 6.6 m/s, with spring being the best performing season.

Resource grids were used via micro-siting to calculate the mean speed, power density, and AEP for both islands. Using the 55 m hub-height Vergnet GEV MP-C WTG, with rotor diameter spanning 32 m, the spatial variation of mean wind speed on Chatham Island ranged from 4.8 m/s to 14.80 m/s, AEP from 3.01 GWh/year to 13.87 GWh/year, and power density from 189 W/m² to 3760 W/m². The wind resource classification for a good portion of the island is Wind Power Class 7, with wind speed and power density greater than 8.7 m/s and 800 W/m², respectively, at 55 m height. For Waiheke Island, the 80 m hub-height Siemens SWT-2.3-82 WTG was used. Mean wind speeds ranged from 5.3 m/s to 8.99 m/s, AEP from 2.93 GWh/year to 9.22 GWh/year, and power density from 167 W/m² to 766 W/m². A sparse portion of the island has a Wind Power Class 6 wind resource. The mean wind speed, the mean annual energy production (AEP), and the mean power density for Chatham and Waiheke Islands are 11.08 m/s, 7.14 m/s, and 11.873 GWh/year, 6.149 GWh/year, 1449 W/m², and 406 W/m², respectively.

Chatham wind farm can produce a total gross AEP of 2.78 GWh/year, total net AEP of 2.73 GWh/year, with a proportional wake loss of 1.85%. SSA shows that the wind farm site is not suitable given that the turbine sites experience mean wind speeds and turbulence higher than the IEC 61400 standard recommendation. The SSA conducted at other potential wind farms on Chatham Island, especially in locations with class 7 wind resource, show that WTGs of 55 m hub height will experience mean wind speeds and turbulence higher than IEC Class IA (the highest WTG classification). WTGs with lower hub heights and specially designed to address challenges with turbulence, even in “complex” terrain, are solutions to explore.

Waiheke Island has a sparse distribution of sites with Class 6 wind resource. The SSA conducted on potential wind farms sited shows that, of the fourteen (14) sites considered, three (3) have very good site feasibility for siting class IIC WTGs.

For a successful wind farm, special consideration needs to be given due to the complexity of the terrain in these sites. Solutions to explore include wind turbine generators with lower hub-heights designed to address the challenges with turbulence intensity.

Footnotes

Acknowledgments

The lead author was the recipient of scholarship funding from the New Zealand Ministry of Foreign Affairs and Trade (NZMFAT). The authors gratefully acknowledge the NIWA and MetraWeather for providing the ground-based wind data measurements from AWSs on Chatham and Waiheke Islands, the Department of Wind Energy of the Denmark Technical University for providing a student license for the wind simulation software package WAsP, WAsP Engineering, and WAT.

ORCID iD

Kunal K. Dayal

Contribution list

Enigbe S. Ocheklie—conceptualization, methodology, software, formal analysis, visualization, investigation, and writing—original draft. Kunal K. Dayal: conceptualization, methodology, software, investigation, writing—original draft, writing—review and editing, supervision. Rajnish N. Sharma: writing—review and editing, supervision.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability

With a valid login, users can access Chatham Island's data from the URL (https://cliflo.niwa.co.nz/) provided. Waiheke Island's data can be purchased from MetraWeather (MetraWeather, 2025).

References

Astrup

Larsen

(1989) WAsP Engineering Flow Model for Wind over Land and Sea.

Bechmann

, et al. (2017) Data Requirements for WAsP, CFD & WRF. DTU Wind Energy E, Vol. 155.

Busby

(2012a) Assessment: Mapping and Measuring the World’s Wind,” in Wind Power - The Industry Grows Up, PennWell, 65–99.

Busby

(2012b) Wind Farms: Developing and Operating Wind Power Plants, in Wind Power - The Industry Grows Up, pp. 101–148.

Chandel

Murthy

KSR

Ramasamy

(2014) Wind resource assessment for decentralised power generation: Case study of a complex hilly terrain in western himalayan region. Sustainable Energy Technologies Assessments 8: 18–33.

Cherry

(1987) Wind energy resource survey of New Zealand : national resource assessment : summary and final report. Auckland, N.Z.: New Zealand Energy Research and Development Committee, University of Auckland 1987.

Dawber

(1988) Wind energy resources in Otago and Southland : final report on Phase II work : the third report of the Otago Division of the Wind Energy Resource Survey of New Zealand. Auckland, N.Z.: New Zealand Energy Resource and Development Committee 1988.

Dawber

Drinkwater

(1996) Windfarm prospects in central otago, New Zealand. Renewable Energy, 9(1–4): 802–805.

Dayal

, 2015. Offshore wind resource assessment, site suitability and Technology selection for Bligh Waters Fiji using WindPRO . MSc Dissertation. Uppsala University.

10.

Dayal

, 2021. Mesoscale wind resource mapping of the small island states of the southwest pacific region . Ph.D. Thesis. The University of Auckland.

11.

Dayal

Bellon

Cater

, et al. (2021b) High-resolution mesoscale wind-resource assessment of Fiji using the weather research and forecasting (WRF) model. Energy 232: 121047.

12.

Dayal

Cater

Kingan

, et al. (2021a) Wind resource assessment and energy potential of selected locations in Fiji. Renewable Energy 172: 219–237.

13.

Dayal

Cater

Kingan

, et al. (2022) An analysis of the 10 MW butoni wind farm in the tropical southwest pacific island of Fiji. Wind Engineering 46: 1264–1280.

14.

Dayal

Cater

Kingan

, et al. (2024) Evaluation of the mesoscale-microscale (WRF-WAsP) coupling methodology for wind resource parameters in Fiji. Energy Exploration & Exploitation 42(4): 1–17.

15.

DTU Wind Energy . (2016) WAsP Engineering 4.0.

16.

DTU Wind Energy . (2019) WAT 4.5 Help.

17.

Edwards

(1990) Wind energy resource studies in New Zealand. Solar Wind Technology 7(1): 9–14.

18.

Edwards

Dawber

and N. Z. E. R. and D. Committee . (1979) Otago wind energy resource survey. Auckland: New Zealand Energy Research and Development Committee 1979.

19.

Edwards

Dawber

(1980) An investigation of the wind energy prospects in the otago region of New Zealand. Journal of Industrial Aerodynamics 5: 281–296.

20.

EMD International A/S . (2019) WindPRO 3.3.

21.

Gipe

(2006) Large Wind Turbines: IEC Wind Turbine Classes,” Wind-works.org, 2006. [Online]. Available: http://www.wind-works.org/cms/index.php?id=85&tx_ttnews%5Btt_news%5D=289&cHash=bfc12bcfc85c00e8d32a3f7efa725743.

22.

GWEC, Global Wind Energy Council. Global Wind Energy Report . (2023) Available at: www.gwec.net. Accessed on March 7, 2025.

23.

Hindsberger

Boys

Ancell

(2012) Perfect partners: Wind power and electric vehicles - A New Zealand case study. In: Sioshansi

(ed) Smart Grid. Cambridge, UK: Academic Press, 453–477.

24.

Kelly

(2011) History and potential of renewable energy development in New Zealand. Renewable and Sustainable Energy Reviews 15(5): 2501–2509.

25.

Landberg

Mortensen

Petersen

(1996) Wind resource assessment and siting : A wider perspective. Water Energy International 53(4): 29–33.

26.

Maslin

Herald

(2019) “Chatham Islands’ isolation makes it a unique and special place,” [Online]. Available: https://www.nzherald.co.nz/the-country/news/article.cfm?c_id=16&objectid=12234069. [Accessed: 27-Jan-2025].

27.

Mason

Page

Williamson

(2010) A 100% renewable electricity generation system for New Zealand utilising hydro, wind, geothermal and biomass resources. Energy Policy 38(8): 3973–3984.

28.

MBIE . (2017) “Unlocking our energy productivity and renewable potential : New Zealand energy efficiency and conservation strategy 2017-2022,” p. 24.

29.

MBIE . (2024) Energy in New Zealand. Energy Build. Trends 20: 58.

30.

McWhirter

(2010) CHATHAM ISLAND WIND GENERATION PROJECT.

31.

MetraWeather . (2025) MetraWeather: Powerful Weather Intelligence. [Online]. Available: https://metraweather.com/about-metraweather. [Accessed: 22-March-2025].

32.

Mortensen

, et al. (2012) Planning and Development of Wind Farms : Wind Resource Assessment and Siting, 3272.

33.

Mortensen

Heahtfield

Rathmann

, et al. (2019) Wind Atlas Analysis and Application Program: WAsP 12 Help Facility. Roskilde, Denmark: Department of Wind Energy, Technical University of Denmark.

34.

Nasab

Kilby

Bakhtiaryfard

(2020) Wind power potential assessment for three locations in New Zealand. Chemical Engineering Transactions 81: 73–78.

35.

New Zealand Wind Energy Association . (2025) New Zealand’s Wind Farms. [Online]. Available: http://www.windenergy.org.nz/wind-energy/nz-windfarms. [Accessed: 26-Jan-2025].

36.

NZ Ministry of Economic Development . (2011) Developing Our Energy Potential: New Zealand Energy Efficiency and Conservation Strategy 2011–2016.

37.

Pincelli

Hinkley

Brent

(2025) Developing onshore wind farms in Aotearoa New Zealand: Carbon and energy footprints. Journal of the Royal Society of New Zealand 55(4): 1005–1027.

38.

Poletti

Staffell

(2021) Understanding New Zealand’s wind resources as a route to 100% renewable electricity. Renewable Energy 170: 449–461.

39.

Pyle

(2016) Case Study on The Development of the wind industry in New Zealand, no. September, 2016.

40.

Rao

(2019) Wind Energy for Power Generation. Switzerland: Springer International Publishing.

41.

REN21 . (2023) Renewables 2023 Global Status Report (Paris: REN21 Secretariat).

42.

Reuther

Thull

J-P

(2011) Feasibility Study of Small and Micro Wind Turbines for Residential Use in New Zealand, no. 30.

43.

Sharma

Ahmed

(2016) Wind energy resource assessment for the Fiji Islands: Kadavu Island and Suva Peninsula. Renewable Energy 89: 168–180.

44.

Tonkin

Beavis

O’Sullivan

(2017) Modelling Wind Flow Over Chatham Island. In: 9th Asia-Pacific Conference on Wind Engineering, Auckland, New Zealand, 3–7 December 2017.

45.

Troen

Lundtang Petersen

(1989) European Wind Atlas. Roskilde, Denmark: Risø National Laboratory.

46.

Underhill

and NBR . (2017) “Owner of mothballed wind turbines on Chatham Islands finally put in liquidation,” [Online]. Available: https://www.nbr.co.nz/article/owner-mothballed-wind-turbines-chatham-islands-finally-put-liquidation-b-206037. [Accessed: 27-Jan-2025].

47.

UNDP, “Goal 7: Affordable and clean energy,” . (2020) [Online]. Available: https://www.undp.org/content/undp/en/home/sustainable-development-goals/goal-7-affordable-and-clean-energy.html. [Accessed: 23-Jan-2025].

48.

United Nations . (2015) Paris Agreement.

49.

Vector . (2019) Vector to help Waiheke Island become the world’sfirst fully electrified island,” [Online]. Available: https://www.vector.co.nz/news/vector-to-help-waiheke. [Accessed: 20-April-2025].

50.

Vergnet Wind Turbines . (2014) “GEV MP C.

51.

Vergnet Wind Turbines . (2020) “Thrust coefficient (Ct) curve.” p. 1.

52.

Waiheke Initiative for Sustainable Energy . (2020) “Waiheke Wind.” [Online]. Available: http://www.waihekewind.co.nz/. [Accessed: 11-May-2020].

53.

WAsP DTU Wind Energy . (2015a) “WAT demo 10 - Wind sector management [Video file].”.

54.

WAsP DTU Wind Energy . (2015b) “WAT demo 5 - IEC site assessment [Video file].”.

55.

Wei Zheng

Yin Li

Jun Xu

(2019) Micro-scale classification of offshore wind energy resource - A case study of the New Zealand. Journal of Cleaner Production 226: 133–141.

56.

White

Wakes

(2014) Permitting best use of wind resource for small wind-turbines in rural New Zealand: A micro-scale CFD examination. Energy for Sustainable Development 21: 1–6.

57.

Zhang

Liu

Zhao

, et al. (2023) Overview of the development and application of wind energy in New Zealand. Energy and Built Environment 4(6): 725–742.

Microscale wind resource and site suitability assessments of Waiheke and Chatham Islands in New Zealand

Abstract

Keywords

Introduction

Theoretical background

The WAsP model

Stability model

Roughness change model

Shelter model

Orographic model

Statistics

WAsP engineering model

Windfarm Assessment Tool (WAT)

Turbine classification

IEC 61400 site assessment criteria

Methodology

Study locations and meteorological data

Software setup and modeling

WAsP

WAsP Engineering

Windfarm Assessment Tool (WAT)

Results and discussion

Temporal variation

Observed mean wind climate

Diurnal variation

Annual variation

Spatial variation

Mean wind speed

Power density

Annual energy production

Wind farm production

Resource classification

Site suitability assessment

Limitations

Conclusions

Footnotes

Acknowledgments

ORCID iD

Contribution list

Funding

Declaration of conflicting interests

Data availability

References