ORBIT: Offshore Wind BOS Installation Tool

• ORBIT is an open-source discrete event simulation tool that models balance-of-system costs and installation logistics for offshore wind projects.
• It decomposes project costs into detailed categories, including hardware, installation, and soft costs, offering scalable analysis for diverse project sizes.
• AIS-derived vessel analytics validate installation time estimates, highlighting variability due to weather, repairs, and other operational delays.

Searching arXiv for ORBIT and related offshore wind installation papers. search_arxiv({"query":"ORBIT offshore renewables balance-of-system installation tool offshore wind", "max_results": 10, "sort_by": "relevance"}) Searching more broadly for offshore wind installation and ORBIT-linked studies. search_arxiv({"query":"offshore wind installation ORBIT AIS machine learning", "max_results": 10, "sort_by": "relevance"}) search_arxiv({"query":"(Tranberg et al., 2019, Bademi et al., 20 Aug 2025)", "max_results": 5, "sort_by": "relevance"}) The Offshore Renewables Balance-of-system Installation Tool (ORBIT) is an open-source, bottom-up design tool that computes balance-of-system (BOS) costs for hypothetical offshore wind energy projects using discrete event simulation. Within offshore wind project analysis, ORBIT functions as a project-level engineering-economic model of BOS hardware, installation logistics, weather downtime, vessels, ports, and project development and “soft” costs; it has also been used as the front end for multiregional economic and greenhouse-gas assessment, while AIS-derived vessel analytics provide an empirical basis for parameterizing and validating installation-time assumptions (Bademi et al., 20 Aug 2025, Tranberg et al., 2019).

1. Definition, scope, and modeling paradigm

ORBIT focuses on balance-of-system and installation aspects of offshore wind projects, not on turbine performance or operations. In the reported usage, BOS includes substructures, array cables, export cables, offshore substations, scour protection, installation logistics, and project development and “soft” costs such as insurance, financing, contingency, commissioning, decommissioning, site assessment and auction, construction planning, and installation planning. This delimitation is important because it distinguishes ORBIT from energy-yield or operations-centered tools: its central object is the cost and schedule structure of offshore wind construction and associated support systems rather than aerodynamic or operational performance (Bademi et al., 20 Aug 2025).

Its core methodological paradigm is discrete event simulation. System actors, such as installation vessels, are modeled as entities executing sequences of tasks including loading components at port, transiting to site, installing foundations, cables, and turbines, waiting during weather downtime, and returning. Events occur in a scheduled sequence, and time and cost accumulate as those events are simulated. In the application to planned U.S. projects, ORBIT was chosen instead of NREL’s JEDI Wind model because it scales costs more accurately with technical design parameters such as capacity, turbine rating, depth, and distance to shore, and because it explicitly models installation logistics.

The tool is documented in Nunemaker et al. (2020), NREL report NREL/TP-5000-77081, and the WISDEM/ORBIT GitHub repository is cited for detailed cost-category calculations and code implementation. A plausible implication is that ORBIT occupies an intermediate position between engineering design tools and techno-economic planning frameworks: it is sufficiently detailed to represent vessel-task logistics, but its outputs are also structured so that they can feed downstream macroeconomic or environmental analysis.

2. Inputs, categories, and internal cost structure

ORBIT is configured from plant-scale, site-scale, and weather inputs. In the five-project application summarized in (Bademi et al., 20 Aug 2025), the key inputs were total project capacity, turbine rating, number of turbines, water depth, distance from site to landfall, mean wind speed, and hourly wind speed and wave height from ERA5 for three years. For “typical” fixed-base offshore wind, default ORBIT assumptions were approximately 23 m water depth and approximately 35 km to shore, overridden where project-specific data existed.

Its output structure follows a detailed decomposition of BOS hardware, installation activities, and soft or development costs. The following categories are explicitly identified:

ORBIT category	Example meaning
Array System	Turbine array cable system design and materials
Export System	Export cable system design and materials
Offshore Substation	Offshore substation structure and equipment
Scour Protection	Rock/armor for foundation scour protection
Substructure	Foundation/base structure construction
Array System Installation	Cable laying vessel operations, port fees
Export System Installation	Export cable installation
Offshore Substation Install	Offshore substation installation
Scour Protection Install	Placement of scour materials
Substructure Installation	Foundation installation
Turbine Installation	Turbine lifting and mechanical completion offshore
Turbines	Manufacture of turbine topside components
Soft costs	Insurance, financing, contingency, commissioning, decommissioning
Project development	Permitting, auctions, engineering plans

Within this structure, each category is computed from equipment quantities determined by plant size and configuration, unit costs for materials, vessels, and labor, time estimates from discrete event simulation, vessel day rates, port fees, and weather delay multipliers. The reported interpretation is therefore explicitly hybrid: quantities and logistics are engineering-determined, while cost formation depends on schedule realization under weather and vessel constraints.

A recurrent misconception is that ORBIT is solely a vessel scheduling model. The documented usage is broader. It produces both installation CAPEX and non-installation categories such as turbine CAPEX, development costs, and annual operation cost for downstream economic payback calculations. Conversely, it is not presented as a full life-cycle model; its main MRIO-linked application centers on the construction and installation phase.

3. Project-level outputs and scaling behavior

In the U.S. application, ORBIT generated total installation cost for each project, defined as BOS plus soft costs and excluding turbines. The reported values were \$296 million for Rhode Island, \$978 million for Maryland, \$2.8 billion for Massachusetts, \$3.0 billion for New York, and \$9.3 billion for Virginia. Across all five projects, total project cost for all components, including turbines, was \$16.3 billion (Bademi et al., 20 Aug 2025).

Those outputs were further interpreted in terms of scaling behavior. The reported result is that the cost of installation per MW of installed capacity decreases with plant upsizing, while turbine manufacturing costs scale linearly with plant capacity because 12 MW turbines were assumed for each project. The ratio of the cost of turbine manufacturing to the cost of installation increases with installed capacity. For smaller projects such as Rhode Island and Maryland, turbine cost is less than installation cost; for larger projects such as Massachusetts, New York, and Virginia, turbine cost dominates.

ORBIT also provides annual operation cost, denoted $C_{\text{op}}$, for economic payback calculations. In the reported framework, ORBIT-derived initial investment cost $C_i$, annual operation cost, a 51% capacity factor, and state-specific retail electricity price are combined to obtain economic payback periods. The published values were 15.2 years for Rhode Island, 11.4 years for Maryland, 5.1 years for Massachusetts, 6.6 years for New York, and 13.6 years for Virginia. These values are properties of the combined ORBIT-based economic framework rather than of ORBIT in isolation.

A plausible implication is that ORBIT’s usefulness increases with project heterogeneity. Because installation cost per MW decreases with plant upsizing while turbine manufacturing scales linearly, the model differentiates between economies of scale in installation and simpler scaling in turbine procurement.

4. Integration with multiregional economic and emissions analysis

The most explicit systems-level use of ORBIT in the provided literature is as the engineering-economic front end to an Environmentally Extended Multiregional Input-Output model of the U.S. economy. ORBIT cost categories are mapped to NAICS sectors and inserted as final demand shocks. For BOS and installation, project costs are assigned to the project-state final demand vector $\Delta y_r$; turbine costs are mapped separately to a turbine-manufacturing final demand vector $\Delta y_i$. The resulting economic impacts are computed with the standard Leontief formulation

$\Delta x = (I - A)^{-1} \Delta y,$

where $A$ is the multiregional direct requirements matrix and $\Delta x$ is total output, including direct and indirect impacts (Bademi et al., 20 Aug 2025).

The same framework attaches greenhouse-gas emissions factors to sector-region outputs. Emissions factors are defined as

$2.8 billion for Massachusetts, \$0

and embodied emissions are computed as

$2.8 billion for Massachusetts, \$1

This coupling allows ORBIT-generated engineering expenditure profiles to be translated into regionally distributed economic and emissions impacts.

The reported aggregate result is \$2.8 billion for Massachusetts, \$216.3 billion in capital investment. Construction and installation emissions were reported as 0.021 million metric tons CO$2.8 billion for Massachusetts, \$3-eq for Rhode Island, 0.085 for Maryland, 0.235 for Massachusetts, 0.295 for New York, and 0.689 for Virginia. Carbon payback periods were reported as 6 months for Rhode Island, 3 months for Maryland, 2 months for Massachusetts, 3 months for New York, and 2 months for Virginia; all projects offset construction-phase emissions in less than a year (Bademi et al., 20 Aug 2025).

The spatial interpretation depends on the multiregional structure rather than on ORBIT alone. Economic spillovers were highlighted for states such as California, New York, Pennsylvania, and Texas, while emissions spillovers were highlighted for Indiana, Ohio, Pennsylvania, and Texas. This suggests that ORBIT’s category-level granularity is sufficiently detailed to support geographically resolved industrial policy analysis when coupled to IO methods.

5. AIS-based empirical calibration and validation of installation modules

A distinct but complementary line of work derives installation-process statistics directly from AIS tracks of jack-up vessels. The stated relevance to ORBIT is explicit: the method provides a concrete, data-driven way to break AIS tracks into operational phases—“harbor,” “transit,” and “installation at turbine”—and to derive per-turbine installation durations and cycle statistics, which is exactly the kind of information ORBIT needs to parameterize and validate its installation modules (Tranberg et al., 2019).

The method uses only GPS positions, latitude and longitude, and does not rely on AIS “Navigational Status,” because that field is manually set and noisy. Tracks are extracted over the installation period for a given jack-up vessel and farm, then K-means clustering is applied to all AIS positions. The number of clusters is chosen as

$2.8 billion for Massachusetts, \$4

where $2.8 billion for Massachusetts, \$5 is the known number of turbines in the farm and $2.8 billion for Massachusetts, \$6 is manually tuned so that “path” points receive their own clusters and do not contaminate turbine clusters. Standard K-means is then used to solve

$2.8 billion for Massachusetts, \$7

with $2.8 billion for Massachusetts, \$8 and $2.8 billion for Massachusetts, \$9.

After clustering, large clusters inside the wind-farm area are taken as turbine locations, and the cluster centroid is the inferred turbine coordinate. Harbor clusters are identified with the same logic. In most farms, reported turbine coverage is at least 94%, and many farms reach 100%. The installation time at turbine $9.3 billion for Virginia. Across all five projects, total project cost for all components, including turbines, was \$0 is defined from visits within a 100 m radius around the centroid,

$9.3 billion for Virginia. Across all five projects, total project cost for all components, including turbines, was \$1

with visit durations

$9.3 billion for Virginia. Across all five projects, total project cost for all components, including turbines, was \$2

and per-turbine installation time

$9.3 billion for Virginia. Across all five projects, total project cost for all components, including turbines, was \$3

For Horns Rev 3, harbor time is obtained from harbor clusters, and transit time is calculated by difference: $9.3 billion for Virginia. Across all five projects, total project cost for all components, including turbines, was \$4

The operational-state interpretation maps naturally to ORBIT. Installation corresponds to the vessel being within 100 m of a turbine cluster; harbor corresponds to presence within harbor clusters while stationary; transit covers all remaining time. In ORBIT terms, harbor maps to at-port tasks such as component loading, crew changes, weather waiting in port, and repairs; transit maps to harbor-to-farm and intra-farm movements; installation maps to on-site tasks such as jack-up and jack-down, tower, nacelle, and rotor lifts, and commissioning. This is not a different model of installation so much as an empirical segmentation that can inform ORBIT’s task durations.

6. Observed installation behavior, uncertainty, and extension paths

Across 16 vessel-farm combinations covering 13 offshore wind farms, the AIS-based analysis reported median installation times of about 20–40 h for many farms, with Galloper reaching approximately 70–80 h medians. Average times were always higher than median because of a few extreme outliers, and the maximum reported installation times reached several hundred hours, including 588.1 h at Hohe See. The stated implication for ORBIT is that installation duration should be treated as a right-skewed distribution with occasional long delays, and that a single global distribution is not appropriate; farm-specific or vessel-specific parameterization is preferable (Tranberg et al., 2019).

The Horns Rev 3 example illustrates the decomposition of vessel time. Over a 4932.5 h operating window for Brave Tern, installation time was 2821 h, total harbor time was 1732 h across 21 harbor segments, and transit time was 379.5 h, corresponding to shares of 57.2%, 35.1%, and 7.7%, respectively. Average harbor stay was 82.5 h, with a maximum of 317.4 h. The reported causes of long harbor times included bad weather, vessel repairs or upgrades, and component shortages. Using wind-speed data from internal Siemens Gamesa databases, the analysis further found that installation time tends to increase with higher average wind speed, that the variance of installation time increases strongly with wind speed, and that the relationship is non-linear rather than simple linear. These observations support weather-sensitive duration models rather than fixed deterministic task times.

The main uncertainties in the AIS methodology are AIS gaps, the manual choice of $9.3 billion for Virginia. Across all five projects, total project cost for all components, including turbines, was \$5, the use of a fixed 100 m radius, and the possibility that long installation times may aggregate multiple segments at the same location, including waiting or repairs rather than only installation. Additional assumptions are that the number of turbines is known, that AIS reporting is relatively frequent, and that one vessel is analyzed at a time. Validation is primarily visual and coverage-based; no direct comparison with detailed project schedules is published (Tranberg et al., 2019).

Limitations of the broader ORBIT-centered framework arise on both the engineering and economic sides. In the MRIO-linked application, only the construction and installation phase is modeled in the main economic and environmental analysis; operations, maintenance, and end-of-life lie outside that main scope, although O&M costs are used for economic payback. Turbine manufacturing is assumed to be domestic in the future, while actual supply chains may remain international. The MRIO analysis is static and therefore does not capture dynamic adaptation of supply chains, learning effects, or price and consumption changes. Proposed extensions include applying the AIS method continuously to create a living empirical library of installation durations by turbine type, water depth, vessel class, and geography; extending clustering and segmentation to multiple vessel types and floating installations; linking ORBIT to dynamic circular-economy tools such as CELAVI; and incorporating scenario-based LCA and multi-criteria decision analysis to evaluate trade-offs among emissions, marine ecosystem impacts, social impacts, and material criticality (Bademi et al., 20 Aug 2025, Tranberg et al., 2019).