Dynamic Simulation and Analysis of the Impact of a planned Windfarm - PDF document

3/30/2011 Dynamic Simulation and Analysis of the Impact of a planned Windfarm on an Isolated Grid Agenda • Introduction • Simulation Model Description • Model Verification • Stability Study • Model Response Analysis • Conclusion 24.02.2011 2 1

3/30/2011 Introduction • Our use of PowerFactory: – Grid connection studies for renewable resource integration. – Assess the need for grid stabilisation. – Solution studies to overcome stability issues. • Purpose of present study: – Windfarm extension on island system. – Create verified simulation model. – Find voltage, frequency or stability issues due to windfarm extension. • Purpose of this paper/presentation: – Show approach for island simulation models – Show the importance of model verification – Show simulation accuracy against real measured data 3 Agenda • Introduction • Simulation Model Description • Model Verification • Stability Study • Model Response Analysis • Conclusion 4 2

3/30/2011 General Information • Island in the Caribbean Sea • Pop (2006): 12,106 • Existing grid - diesel only Power Station – 6 heavy fuel oil generators – Power range 0.9-2.7 MW • Max. Demand: 8 MW 5 General Information • Renewable Integration – 8 Wind Turbines – No grid stabilisation • Model input data – Wind speed and turbulence factor – Load data • System measurements recorded after commissioning in windfarm substation and at generator terminals 6 3

3/30/2011 Grid model description • Grid model Windfarm – Sub-model power station WTG ~ M Power Station – Sub-model windfarm – Sub-model grid Generator Consumer Group 2 ~ • Dynamic models for: G TF1 – Generator Line(3) – Wind turbine Line Line(2) Line(4) – Wind profile Line(1) • Model inputs – Wind speed for WTG – Static consumer load s Consumer Group 1 7 Generator model description • Dynamic model for the generator – Droop functionality – Speed and voltage control – Engine model for turbo-lag effects Qact u_act 0 ve V-Control 0 ElmV* u_set Droop-v 1 0 * 1 Generator 2 ElmSym* fref Droop-f 0 3 * alpha n-Control 0 4 Elmn* pt Diesel Engine 1 1 ElmD* 1 n_act f_act Pact 8 4

3/30/2011 WTG model description • Dynamic model for the wind turbine – Pitch-controller n_act – Rotor P_act – Damper (Shaft) 0 P_set wa_gen Power Set Point – Gearbox 1 ElmP* – ASM 0 0 Gearbox Pitch ElmG* 2 1 1 ElmP* n_set Speed Set Point 3 Elmn* alpha 0 Ma pt 0 Damper 0 On-Off Rotor ElmD* 4 1 0 ElmOn* ElmR* M_rotor ASM 1 * 1 wind_speed 1 2 Wind-Turbulence w_rotor ElmDsl 9 Agenda • Introduction • Simulation Model Description • Model Verification • Stability Study • Model Response Analysis • Conclusion 10 5

3/30/2011 Model verification • Generator model verification – On site step load testing – High resolution data recorder – Different load steps (Generator behaves different) – Parameter adjustment in PowerFactory model • Wind turbine verification against field data DIgSILENT 60.50 60.25 60.00 59.75 59.50 59.25 30.00 34.00 38.00 42.00 46.00 [s] 50.00 Case3E_Gen7: Frequency in Hz GEN7: Simulated Frequency in Hz 11 Agenda • Introduction • Simulation Model Description • Model Verification • Stability Study • Model Response Analysis • Conclusion 12 6

3/30/2011 Stability study • System construction with verified models • Simulation setup: – WTG limited to max power output – Wind model determines wind speed for WTG – Frequency and voltage control by 1 generator – Other generators in power set point mode • Simulation system allocations: – Simulation duration 360 seconds – Mean wind speed 8m/s with 12% turbulence – No wind park spatial dispersion – Load is constant 13 Simulation results • Initial model before DIgSILENT 64.00 commissioning 62.00 – Frequency variation 60.00 4Hz – Average variation in 58.00 Power 400kW 56.00 100.0 172.0 244.0 316.0 388.0 [s] 460.0 – Timeframe 360sec GEN7: Electrical Frequency in Hz 2.40 2.00 1.60 1.20 0.80 0.40 100.0 172.0 244.0 316.0 388.0 [s] 460.0 GEN7: Active Power in MW 14 7

3/30/2011 Real System Measurements • Frequency variation 1Hz • Average variation in Power 200kW • Timeframe 360sec 15 Agenda • Introduction • Simulation Model Description • Model Verification • Stability Study • Model Response Analysis • Conclusion 16 8

3/30/2011 Analysis of model response • Generator step response – Poor measurements for initial step response tests; larger steps were needed to see the full (nonlinear) response of the engine – Resultant model frequency stiffness was lower than reality; this gave the false indication of high frequency variation • Wind turbulence factor – Turbulence factor very difficult to measure – High turbulence factor gave high system frequency variation • Wind park spatial dispersion – Wind park spatial dispersion factor ignored in initial simulations – spatial dispersion factor reduces frequency variation 17 After commissioning • Simulation response after commissioning and adjustment – Frequency variation DIgSILENT 61.50 1Hz 61.00 – Average variation in 60.50 Power 200kW 60.00 – Timeframe 360sec 59.50 59.00 100.0 172.0 244.0 316.0 388.0 [s] 460.0 GEN7: Electrical Frequency in Hz 2.40 2.00 1.60 1.20 0.80 0.40 100.0 172.0 244.0 316.0 388.0 [s] 460.0 GEN7: Total Active Power in MW 18 9

3/30/2011 Model verification results Model Wind Frequency Turbulence deviation Real System Unknown 1.0 Hz(peak) Initial modelling of 12% 4.0 Hz system Model after first step 12% 3.0 Hz load tests Model after 12% 1.5 Hz commissioning Different Turbulence 6% 1.0 Hz Wind park model with 12% 1.0 Hz spatial dispersion 19 Agenda • Introduction • Simulation Model Description • Model Verification • Stability Study • Model Response Analysis • Conclusion 20 10

3/30/2011 Conclusion • Three main aspects of the paper: – Dynamic model development • Correct design of generator and WTG models – Parameterization of models • Generator model parameters wrong due to measurements – Input data for the simulations • Modeled wind profile correct but turbulence undefined • Wind park spatial dispersion • Conclusions – Wide range of generator step tests required for proper model verification – High resolution wind data required for wind profile – Consider other factors such as WTG spatial dispersion, load dynamics, etc. 21 Head Office Europe Office Darwin Business Park Wissenschaftspark Trier Export Drive Am Wissenschaftspark 29 Berrimah 0828, N.T. 54296 Trier Australia Germany mail@pcorp.com.au mail@powercorp.eu +61889470933 +4917683234822 www.pcorp.com.au www.powercorp.eu 22 11