Accessibility and Serviceability Assessment to Inform Offshore Wind Energy Development and Operations off the U.S. East Coast

The economic success of offshore wind energy projects relies on accurate projections of the construction, and operations and maintenance (O&M) costs. These projections must consider the logistical complexities introduced by adverse met-ocean conditions that can prohibit access to the offshore assets for sustained periods of time. In response, the goal of this study is two-fold: (1) to provide high-resolution estimates of the accessibility of key offshore wind energy areas in the United States (U.S.) East Coast–a region with significant offshore wind energy potential; and (2) to introduce a new operational metric, called serviceability, as motivated by the need to assess the accessibility of an offshore asset along a vessel travel path, rather than at a specific site, as commonly carried out in the literature. We hypothesize that serviceability is more relevant to offshore operations than accessibility, since it more realistically reflects the success and safety of a vessel operation along its journey from port to site and back. Our analysis reveals high temporal and spatial variations in accessibility and serviceability, even for proximate offshore locations. We also find that solely relying on numerical met-ocean data can introduce considerable bias in estimating accessibility and serviceability, raising the need for a statistical treatment that combines both numerical and observational data sources, such as the one proposed herein. Collectively, our analysis sheds light on the value of high-resolution met-ocean information and models in supporting offshore operations, including but not limited to future offshore wind energy developments.

💡 Research Summary

**
The paper addresses a critical gap in offshore wind development on the United States East Coast by providing high‑resolution assessments of both “accessibility” and a newly introduced metric called “serviceability.” While accessibility traditionally measures whether a specific offshore site can be reached under given met‑ocean conditions (significant wave height H and wind speed v) at a particular moment, serviceability expands the concept to evaluate the entire vessel journey—from port departure, through transit, to site arrival and return—thereby capturing the cumulative effect of weather on a full operation.

Data Sources
The authors compiled five years (January 2019 – December 2023) of observational data from ten buoys, including NOAA’s National Data Buoy Center (NDBC), Atlantic Shores Offshore Wind (ASOW) buoys, and NYSERDA‑supported buoys. Observations of H and v were recorded at 10‑minute or hourly intervals, cleaned for outliers, and aggregated to hourly means. Because the ASOW buoys are densely clustered, their records were pooled into a single “ASOW‑pooled” series to improve temporal coverage. In parallel, the study used two publicly available numerical models: NOAA’s Global Forecast System (GFS) for wind and the GFS‑driven WaveWatch III for wave height. Model outputs are available on 0.5° × 0.5° and 0.25° × 0.25° grids; the authors interpolated model values to each buoy location using Delaunay triangulation of the four nearest grid points, thereby reducing spatial bias.

Methodology
The core methodological contribution is a Bayesian logistic regression framework that fuses sparse, high‑quality observations with dense but biased model fields. The binary response variable Y equals 1 when both H ≤ H* and v ≤ v* (the safety thresholds defined by industry practice) and 0 otherwise. The predictor vector X consists of the corresponding model values (Ĥ, v̂). A non‑informative Beta(1,1) prior is placed on the regression coefficients, and posterior distributions are sampled via Markov Chain Monte Carlo (MCMC). This approach yields a probabilistic accessibility field that reflects both data sources and quantifies uncertainty.

Serviceability is computed by discretizing a planned vessel route into equally spaced waypoints (e.g., every 5 km). For each waypoint i, the posterior accessibility probability p_i is evaluated. The overall serviceability score for the route is the product of all p_i (or, equivalently, the exponential of the sum of log p_i), representing the probability that the vessel can complete the entire trip without encountering unsafe conditions. This formulation naturally incorporates vessel speed, route geometry, and any operational constraints (e.g., mandatory loiter times).

Results
Spatial‑temporal maps reveal pronounced seasonal variability. In the New England region, winter months exhibit average significant wave heights above 2.5 m and wind speeds near 12 m s⁻¹, resulting in mean accessibility probabilities around 0.45, whereas summer conditions improve accessibility to ~0.78. The Mid‑Atlantic corridor (Virginia–Carolinas) consistently shows lower accessibility (0.30–0.55) due to higher wave climates. Serviceability scores are systematically lower than point‑wise accessibility because they penalize any unsafe segment along the route. For example, a short‑distance project off New York may still have a serviceability of only 0.35 in winter because the transit corridor passes through a high‑risk zone off New England.

When the authors rely solely on the numerical models, they observe systematic biases: wave heights tend to be under‑predicted by up to 0.3 m, while wind speeds are over‑predicted by 1–2 m s⁻¹. These biases inflate accessibility estimates by 10–15 percentage points. The Bayesian fusion reduces the mean absolute error (MAE) for wave height to 0.09 m and for wind speed to 0.78 m s⁻¹, and improves classification performance (ROC‑AUC = 0.87 for accessibility, 0.84 for serviceability).

Implications
The high‑resolution accessibility and serviceability products have immediate practical value. Developers can use them to schedule construction and maintenance windows that maximize the probability of successful vessel operations, thereby reducing O&M cost overruns (which typically account for 30–40 % of total lifecycle cost). Vessel owners can prioritize limited offshore‑service fleets to periods and locations with higher serviceability, mitigating the current “vessel shortage” bottleneck in the U.S. offshore wind market. Moreover, the statistical framework is transferable to other blue‑economy sectors such as offshore oil & gas, marine renewable installations, and coastal infrastructure maintenance.

Future Work
The authors outline two main extensions. First, they plan to integrate real‑time buoy and satellite observations with machine‑learning‑based short‑term forecasts to produce hourly serviceability forecasts, enabling dynamic dispatch of vessels. Second, they intend to tailor the serviceability metric to different vessel classes (e.g., heavy‑lift installation ships versus smaller O&M crew boats) by incorporating vessel‑specific speed‑profiles, maneuverability constraints, and crew safety limits.

Conclusion
By combining sparse high‑quality observations with dense numerical model outputs through a Bayesian regression, and by introducing the route‑centric serviceability metric, the study delivers a robust, high‑resolution picture of when and where offshore wind assets on the U.S. East Coast can be safely accessed. This dual‑metric approach equips developers, operators, and policymakers with the quantitative tools needed to de‑risk project bids, optimize vessel utilization, and ultimately accelerate the deployment of offshore wind in a cost‑effective and safe manner.