Digital Twins as Decision Support Tools

Introduction to Digitalization in Wind Energy

Digitalization is transforming the wind energy sector by enabling advanced asset optimization and decision-making through the use of digital twins. A digital twin is a virtual representation of a physical asset, process, or system that is used to simulate, predict, and optimize performance.

• Asset Optimization: Researchers and asset owners use digital twins to optimize wind farm investments and operations.

• Decision Support: Digital twins provide a platform for evaluating investment decisions over the asset's lifecycle, considering both familiar and new market scenarios.

• Reliability Engineering: The integration of digital data streams has improved the accuracy of reliability predictions and risk quantification.

Key Concepts and Definitions

• Digital Twin: A digital replica of a physical asset, used for simulation and analysis.

• Asset Optimization: The process of maximizing the value and performance of an asset over its lifecycle.

• Decision Support Platform: A system that integrates data, models, and analytics to assist in making informed investment and operational decisions.

• Net Present Value (NPV): The present value of cash inflows minus the present value of cash outflows over a period of time.

• Real Options Valuation (ROV): A method for valuing the flexibility to make future investment decisions.

• Annual Energy Production (AEP): The expected yearly energy output of a wind turbine or wind farm.

• Power Sale Price (PSP): The price at which generated electricity is sold, typically in $/MWh.

• Remaining Useful Life (RUL): The estimated time a component or asset will continue to function before replacement is needed.

Wind Farm Asset Optimization Framework

Framework Overview

The asset optimization framework for wind farms integrates multiple data sources and models to support investment decisions. The framework considers market data, reliability data, and operational data to evaluate different scenarios.

• Data Inputs: SCADA data, aftermarket sensor data, forecasted loading, maintenance records, published reliability data, OEM predictions.

• Rules: Foundation, transformer, tower, yaw motor/drive, nacelle, generator, gearbox, rotor blades.

• Market Data and Simulations: Monte Carlo simulations, market data, published data, industry methodology.

• Decision Support: Real-options valuation, scenario testing.

Model Flowchart

The model flowchart outlines the decision-making process, from data collection to scenario analysis and final investment recommendations. It incorporates user-defined inputs, simulation of future market cases, and live updates of key metrics such as NPV and ROV.

Strategic Decision Support Platform Qualities

Platform Requirements

• Modularity: The platform must handle diverse data types, including reliability prognostics, market futures, and non-financial benefits.

• Accessibility: Standardized input forms and transparent methodologies enable collaboration and reproducibility.

• Visualization: Tools like Discounted Cash Flow (DCF) analyses provide clear, time-sensitive visualizations of asset value.

Fitting Standardized Elements Together

Standardization allows for the integration of various data streams into a unified cash flow model, facilitating comprehensive investment analysis.

• Revenue Opportunities: Enviro-credit valuation, hydrogen futures.

• Opportunities: Blade rules, curtailment reallocation.

• Valuation Methods: Net Present Value, Real Options Valuation.

Simple Case Study: Wind Farm Decision Support

The Setup

A 20-year-old wind farm considers the next 10 years of its service life under various market and operational scenarios.

Current Market Condition

• Case 0: Current PSP and AEP using existing curtailment rates and energy production.

Future Market Changes

• Case 1: New PSP in Year X with P(x) (e.g., new plant opening, 65% certainty, PSP = $125/MWh).

• Case 2: AEP Increase from Ancillary Market in Year X with P(x) (e.g., ancillary market opens, AEP increases by 10%, 65% certainty).

• Case 3: Curtailment Reallocation in Year X with P(x) (e.g., ammonia/hydrogen production from curtailed energy, 50% certainty).

Foundational Variables

• Capacity Factor (CF): 30%

• Installed Capital Cost (ICC): $2,400,000/MW

• Nameplate Capacity: 2.3 MW

• Operating Expenses (OpEx): $31,000/MW

• PSP (original): $100/MWh

• Maintenance Schedule: AEP is normally distributed (SD of 10% per year), simulated via Monte Carlo (10,000 simulations).

• Variance: 32% (tech. + hoep + anc.)

• Discount Rate: 6%

Component RUL and Cost Table

Case Study Results

Case 0: Current AEP and PSP (Base Case)

• Unimodal normal distribution of NPV over 10 years.

• Percentile lines (10th, 50th, 90th) indicate risk and return.

• Real-Options Value (ROV) calculated.

• Recommendation: "Keep Asset" based on NPV and ROV/NPV ratio.

Case 1: Current AEP and New PSP

• Bimodal distribution due to PSP variability.

• Percentile lines and ROV calculated.

• Recommendation: "Keep Asset" based on NPV and ROV/NPV ratio.

Case 2: New AEP and Current PSP

• Unimodal normal distribution with increased AEP.

• NPV and ROV slightly higher than Case 0.

• Recommendation: "Keep Asset" based on NPV and ROV/NPV ratio.

Case 3: New AEP and New PSP

• Bimodal distribution due to both AEP and PSP changes.

• NPV and ROV calculated.

• Recommendation: "Keep Asset" based on NPV and ROV/NPV ratio.

Dashboard and User Inputs

The interactive dashboard allows users to input variables, run simulations, and visualize results in real time. Key user inputs include PSP, AEP, OpEx, ICC, RUL, and maintenance schedules.

Sample Dashboard Input Table

Next Steps

• Incorporate more detailed maintenance schedules.

• Use actual RULs and real AEP data for improved accuracy.

Key Equations

• Net Present Value (NPV): where is the net cash flow at time , is the discount rate, and is the number of periods.

• Real Options Valuation (ROV):

• Annual Energy Production (AEP): where is the rated power (MW), is the capacity factor, and 8760 is the number of hours in a year.

Summary Table: Case Study Scenarios

Conclusion

Digital twins and advanced decision support tools are essential for optimizing wind farm assets in a rapidly evolving energy market. By integrating real-time data, reliability engineering, and financial modeling, asset owners can make informed decisions that maximize value and minimize risk.

Additional info: This summary expands on the slides by providing definitions, formulas, and context for key concepts in wind energy asset optimization and digitalization.