The offshore wind industry is growing faster than ever and is poised to be one of the post-Covid industrial growth machines. Consequently, the need to faster analyse more projects and more complex scenarios have risen, as has the need for much better decision support tools in the increasing competitive landscape
Written by: Ole-Erik Vestøl Endrerud, PhD
Now that the industry has achieved a substantial cost reduction to make offshore wind a genuinely competitive energy source, projects demand more detailed and accurate analysis tools to understand increasingly complex scenarios, estimate risk and improve financial viability. The emergence of floating wind as a viable alternative to bottom fixed wind turbines and feeder-based installation concepts in the US are also two special cases that Shoreline’s Design software has proven agent-based modeling can handle with ease.
This post discusses four reasons why you can use agent-based simulation to do precisely that. Initially developed by the Santa Fe Institute in California to study complex systems, it has become widely used to look at anything from ecology to operation and maintenance in offshore wind.
So why are agent-based simulations so valuable?
- Precise estimates from accurate modeling means better performance and lower costs
- One unified system means a single model will provide every required output
- Complex scenarios like resource sharing, floating wind and feeder-based installation, can be handled with ease
- Faster and less costly analysis means real-time decision making and on the fly adjustments
And now for a little more detail…
#1 Precise estimates from accurate modeling means better performance and lower costs
As an industry, we’re incredibly ambitious in driving costs down further, and innovative in thinking of new installation and operational concepts to do so. But that doesn’t come without challenges. These new concepts, such as using service trains to maintain several wind farms in the same region and more complex supply chains and logistics for installation projects, require more advanced models to optimise costs, contracts, and projects.
Agent-based modeling was created for just this purpose. Agent-based modeling models each act in the system independently, for example, how a vessel transports people to an asset, how a technician performs maintenance activities on the asset, and how a jack-up performs lifting operations. In a mathematical or analytical model, you would model the system, not the actors. For example, an installation estimation using a mathematical model could assume the rate of turbines installed per month as input and add some weather downtime. An agent-based model would model how components are pre-assembled, loaded out, transported, installed, and commissioned and simulate all installations as they would occur in real life. These simulations would then be used to estimate the installation rate, hence, giving you a significantly more accurate and precise estimate of an installation campaign.
Instead of modeling how probable it is that something happens, you model how it happens. For example, you can simulate how much time you get to work on a turbine based on the weather, the vessel’s weather limitations, the HSE limitation on the turbine, and the technician’s skill level, instead of just using a random number generator to estimate the expected repair time for a task.
#2 One unified system means a single model will provide every required output
Traditionally we have several models at work to look at different outputs for a project. There is a weather downtime estimation model to estimate monthly workability, a reliability model to estimate availability and downtime, a project management tool to look at installation master schedules, and so on.
With an agent-based modeling tool like Shoreline’s Design software, you can estimate all of this in the same simulation, where you also capture all the interdependencies between these factors. For example, the knock-on effects of the production rate of components in Asia for your project in Europe, the bottleneck of a slow turbine installation campaign, or the need for fewer technicians for summer campaigns when you share an SOV with other wind farms in the area.
Because agent-based models are so realistic it is easy to measure fuel consumption, flying hours, downtime per turbine, the number of turbine transfers, the exact split of downtime of an installation vessel between weather and waiting on other tasks, to mention a few examples.
#3 Complex scenarios like resource sharing, floating wind and feeder-based installation, can be handled with ease
It isn’t difficult to understand that when you start looking into more complex scenarios such as feeder-based installation, floating turbine maintenance, or resource sharing, the small handy spreadsheet model starts to become rather large and unmanageable while assumptions become increasingly less clear. Adding such new methods to an agent-based model is easy as nothing else in the model changes, you just have to add a new type of vessel – say a feeder barge or a tug – and model how that should behave and how the task it performs should be carried out. Shoreline’s Design software shows how an agent-based model can add everything in a construction project from fabrication, component transport, pre-assembly, loadout, installation, commissioning, and testing and precisely simulate and estimate how such a project will perform.
#4 Faster and less costly analysis means real-time decision making and on the fly adjustments
With the incredible growth the offshore wind industry is experiencing, you have to analyse more projects than ever. Agent-based models that are built well run incredibly fast, and a life cycle analysis of a project can be completed in a matter of hours from start to output.
This speed comes with great benefits. Because it takes so little time to run a scenario, you can significantly increase the number of what-if analyses you can do simultaneously. Then, when you present your results to management, you can be sure you’ve looked at all the scenarios, including edge cases, or even make changes live while presenting. A simulation run takes between 5 and 30 minutes to complete, which means you can enable real-time decision-making in meetings where you can quickly change a scenario and rerun in minutes to see how the outputs change.
Looking ahead, the need for fast and accurate tools that can handle complex scenarios will only increase as the offshore wind industry grows further, with agent-based modeling providing a faster, proven and capable method to use for these growing set of requirements. And to keep up simulation tools such as Shoreline Design keep evolving to cover the need by continuously developing new modeling capabilities for offshore wind project development, no matter how complex it becomes.
Interested in using agent-based modeling to optimise your day to day work scheduled on site? Check out Shoreline Execution our execution platform for construction management and asset management
Shoreline Execution the integrated platform for construction management and asset management