Abhinav Datla is a student at the University of Edinburgh, reading for an MPhys in Physics. He joined the Analysis & Insights team for eight weeks over the summer, as part of a research placement organised by the university. This is the second placement in the Analysis & Insights team, which hosted Frances McGinley last year. Here, Abhinav discusses the new direction for the simulation tool, COMPASS:
The COMPASS operations & maintenance (O&M) simulation tool is approaching three years of development and has been used in a variety of applications across the wind industry. It has been effective in accurately imitating hypothetical wind farms and breaking down different cost estimates, utilization statistics, OPEX, and KPIs for a range of vessel types. In the past eight weeks, I have had the privilege of working as an intern in the Analysis & Insights (A&I) team at ORE Catapult, to refine COMPASS in such a way that increases the reliability, provides useful visual outputs, and creates an accessible user experience.
Overall, the outcome is a faster, more accurate and more marketable O&M simulation software.
Simulating with COMPASS
To briefly summarise the current state of COMPASS, there are several different interacting simulation objects incorporated that each have specific parameters attached to them. These include maintenance vessels, cables, generators, substations, ports, and personnel. These are coupled with additional inputs such as the weather/sea-state data, farm layout, and different failure/maintenance events. The events can take on different forms and apply to substations, cables, generators, and/or other miscellaneous assets.
COMPASS distinguishes planned and unplanned repairs. Planned repairs are scheduled at specific intervals, typically monthly, allowing for proactive management of wind farm components. Unplanned repairs modelled using the Monte Carlo (random-sampling) simulation, account for both fixed and variable failure rates, providing realistic and dynamic failure scenarios. These repair events can be merged with the actual maintenance activities, offering a complete view of the maintenance processes and ensuring that all aspects of wind farm operation are managed.
The COMPASS interface primarily utilises Excel for input/output. Users select the farm lifetime (the length of time you plan on having the software simulate) and the number of runs. At the end of a simulation, it outputs a wide range of KPIs for the wind turbines and vessels including time availability, production-based availability (energy availability), and cost breakdown.
Visualisation and Interactivity
COMPASS is a Monte Carlo-based simulation tool, meaning it is necessary to have multiple runs (~50) to obtain a converging value, with an error that falls within an acceptable statistical tolerance. Achieving this efficiently can pose challenges, as each simulation run can range in length from seconds to hours, depending on the size and complexity of the farm. It’s essential to have an effective conversion/visualisation of the data, which mitigates the challenges and efficiently produces a useful result.
One such example is a histogram, which would generate a figure that plots the frequency of each parameter’s output values. In the example implementation below, three different survey vessels are compared by their transit total, and the dotted line marks their mean. A drop-down menu allows the user to select the displayed parameter. As the number of runs increases, the error on the output mean reduces, allowing for accurate estimates.
Figure 1: Example implementation of histogram displaying convergent mean values for n(~100) number of runs.
When determining where to modify the program, it is challenging to recognise whether a given factor has a significant impact on the overall output. Different considerations are brought up by users and then evaluated for their effect on the result and simulation runtime.
For example, the O&M vessels each have a power consumption value attached to them, subject to change depending on different weather conditions. Headwinds and high sea-state during transit may increase the power consumption for example, which must be accounted for. Visualizing the impact of this change is done through a radar plot, in which different values can be plotted before and after the change. This makes it clear what values of interest are affected and verifies what they physically represent. This modification, for example, is expected to have a larger “Money Hire Total” and shorter “Total Wait on Weather”. This plotting method is most effective when used after the histogram so that a converged mean value can be used and a proper range can scale the values within the plotting range.
Figure 2: Computer illustration of headwind on O&M vessel, showing airflow patterns over the ship and its resistance (or assistance depending on direction) due to wind.
Figure 3: Example implementation of radar chart before (blue) and after (yellow) accounting for the effect of headwind on power consumption for O&M vessels.
Further Development and Release
COMPASS is continually evolving to meet the dynamic needs of the wind industry. As it approaches its fourth year of development, the focus is on enhancing its capabilities and expanding its applications.
Currently, the tool is run by the COMPASS team and is available on request. However, we hope to transition to a widespread product, where different release methods are possible and can be tailored to meet different user needs.
The tool can be accessed as a standalone application, via a subscription-based Azure cloud-hosted application or as a subscription-based/one-time purchase licensed software distribution. For those preferring a web-hosted application, COMPASS can be deployed through Azure App Service, utilizing web jobs or API integrations for backend operations hosted on ORE Catapult servers.
The tool is utilised across the wind industry. It is currently used in the Newhaven Rampion Biogas project, to analyse the Rampion 2 wind farm, focusing on charging locations and the impact of electric or hydrogen vessels on operational costs. Additionally, it is being employed by our partners in Korea, to run simulations on a generic wind farm, adjusting parameters through a web API/GUI. COMPASS is revolutionising O&M strategies in the renewable energy sector, ensuring efficient and sustainable operations for wind farms worldwide.
Overall, I am proud to have been part of the A&I and COMPASS teams. I am looking forward to applying these skills in my career and being a part of the future of renewable energy.