This page contains a Flash digital edition of a book.
creating wind turbines Analysis results with


Google Earth For the analysis of wind farms, Ansys has developed WindModeller, a set of automated tools that automates the process of setting up computational fluid dynamics (CFD) simulations of wind-turbine farms. It simplifies the process of importing map data (terrain), meshing, setting up loads (different wind directions), solving and post- processing, and reporting. As for the last step, it is even possible to use Google Earth to display results in a more meaningful way. WindModeller is not an official Ansys


product, explains its champion Christiane Montavon, but instead is a set of scripts and binaries that can be operated from a simple user interface to make this type of modelling easier for those not steeped in CFD techniques. It does not itself require a license, but can only be used in conjunction with Ansys’ CFX and Fluent flow solvers. The tool has been made available to selected customers, and several new customers developing wind farms have licensed the CFD products with the sole intention of using WindModeller. Montavon relates how the tool has been used to simulating the flow conditions for sites such as the isolated Bolund Hill in Denmark, the more complex terrain of the Nant y Moch site in central Wales and the Black Law wind farm in central Scotland.


Into the grid Besides using Matlab and related tools to study individual components of a wind turbine, engineers also use it for other system studies. For instance, before bringing new wind farms online, Hydro-Québec conducts extensive simulations to plan their integration into the grid, forecast power output and ensure safe, reliable operation. The company


same large power system. Further, using our stability simulation tools we were able to determine more accurately the amount of dynamic shunt compensation needed. When our simulations show that we can reliably operate with just one less SVC, it saves Hydro-Québec millions. ‘Finally, with traditional tools we could


A wind rose, here generated with Matlab, uses a polar coordinate system to plot the frequency of winds over a long time period


wind farm to validate an aggregation method for modelling wind farms, and then simulate their interaction with the power network. Hydro-Québec engineers modelled


individual turbines and entire wind farms using tools from The MathWorks such as Matlab, Simulink and SimPowerSystems (a product they themselves continue to develop); generated code from the models with Real-Time Workshop; and used this code in their multiprocessor environment to evaluate wind farm performance in the context of the power system as a whole. They also built wind turbine models in Simulink using power electronics blocks from SimPowerSystems and also built Simulink models of generator control systems. To study stability, Hydro-Québec


simulated the mechanics of a turbine using a two-mass system model in Simulink that accounted for the pitch of the blade and torsional effects on the drivetrain. They then assembled a Simulink model of an entire wind farm consisting of 73 individual turbine models and the collector network that links them. Using Real-Time


FOR WIND FARMS, BUSINESS PLANNING PRESENTS A CHALLENGE BECAUSE A WIND FARM’S PRODUCTION CHANGES BASED ON LOCAL WIND SPEEDS


must determine how much equipment, such as static VAR compensators (SVCs), will be needed when new wind turbines are connected to the electrical grid. ‘Without accurate models, we risk installing millions of dollars’ worth of unnecessary equipment or not having the equipment we need to meet our reliability and production goals,’ says Denis Laurin, manager of technological innovation at Hydro-Québec TransÉnergie. Hydro-Québec needed to model the control system and power output of a single turbine, simulate the 70 to 100 turbines of a single


34 SCIENTIFIC COMPUTING WORLD


Workshop, the team generated C code from their Simulink and SimPowerSystems models, which they ran in Hydro-Québec’s Hypersim simulation environment on a 32-processor supercomputer. ‘Previously, completing a straightforward


simulation of a few seconds for turbines connected to a large power system took hours,’ says Laurin. ‘Now we can complete complex electromagnetic transient (EMT) simulations for entire wind farms in seconds, and we obtained real-time simulation speed for aggregate wind farms connected to the


not study the interaction between series capacitance and the wind power plant. The EMT aggregate wind farm models enabled us to perform large-scale power systems studies such as the interaction between series compensation and wind farms.’


Studying dollars and cents Another aspect of wind-power facilities, one that doesn’t come immediately to mind as being a candidate for modelling and simulation, is the financial end. Utilities that run coal, natural gas, and oil power plants can control production and predict future revenue. For wind farms, however, business planning presents a challenge because a wind farm’s production changes based on local wind speeds. To generate accurate revenue forecasts and


revenue-at-risk projections, one wind farm operator has combined production estimates for all the wind farms in its portfolio with forecasts of power prices on the futures market. The engineers had to link price forecasts to volumetric forecasts for wind- generated power while cross-correlating price and wind levels across multiple geographically dispersed locations. The company needed an automated forecasting and risk-management solution that was reliable and scalable and could be deployed easily within the existing IT infrastructure. Using Matlab, analysts developed an


automated risk-forecasting system that factors in historical data, current prices, and estimates from expert analysts. The company previously had a network of 15 spreadsheets, some with as many as 500,000 rows; the system would crash, it wasn’t scalable, and each new run required hours of manual work. The programmers started with the Database Toolbox to read data from multiple SQL databases, including internal long-term price forecasts, third-party forecasts, historical prices and daily forwards contracts. They next developed algorithms in Matlab to analyse this data to produce monthly price forecasts for the next several years across all the operator’s wind farm sites. They then used the Matlab Compiler to deploy a version of the program that runs automatically each morning and stores its forecast results in the database.


www.scientific-computing.com


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44