1. www.risoe.dk Hybrid System Performance Evaluation Henrik Bindner, Tom Cronin, Per Lundsager, Oliver Gehrke Risø National Laboratory, Roskilde, Denmark EWEC, 2nd March 2006, Athens

2. www.risoe.dk Hybrid System Performance Evaluation Hybrid Systems – the Problem How will SYSTEM 1 perform at SITE 1? Is SYSTEM 1 better than SYSTEM 2 at SITE 2? Can you trust the answers? Key Issue: Develop method for system performance assessment Main ingredient: System modelling, especially supervisory controller modelling

3. www.risoe.dk Hybrid System Performance Evaluation This Work – the Questions The presentation deals with the following issues of modelling of hybrid system:  The optimum timestep to adequately take account of the wind variability.  Sensitivity of various performance criteria to the choice of timestep.  The nature of controller approximations necessary for longer timesteps.

4. www.risoe.dk Hybrid System Performance Evaluation This Work – the Approach Model simple system with 2 diesel gensets, 3 wind turbines and 3 loads Specify performance criteria Investigate performance for 2 different controllers and for 3 different simulation time steps

5. www.risoe.dk Hybrid System Performance Evaluation IPSYS – an explanation IPSYS is a flexible framework for simulation of integrated energy systems IPSYS is a dispatch model - modelling timesteps are in the range 10s-60min IPSYS explicitly models the power system and does load sharing between the frequency controlling units IPSYS is multi-domain – it models the electrical, mechanical, thermal, water or other domains IPSYS is very flexible in modelling configurations and supervisory controllers

6. www.risoe.dk Hybrid System Performance Evaluation Layout of the Simulated Hybrid System G Combined Load 1 Combined Load 2 Combined Load 3 WTG Dump Load G

7. www.risoe.dk Hybrid System Performance Evaluation Main System Details 2 x diesel generators: 780kW Generation voltage: 400V Step-up transformer: 1.7MVA Distribution voltage: 22kV 3 x 300kW wind turbines No-load reactive power compensation: 250kVA 900kW dump load 3 x geographically dispersed loads

8. www.risoe.dk Hybrid System Performance Evaluation Load Profiles

9. www.risoe.dk Hybrid System Performance Evaluation Droop Curve for Diesel Generators and Dump Load

10. www.risoe.dk Hybrid System Performance Evaluation Diesel Generator Controls ON: 75% Pr = 570kW OFF: 25% Pr = 195kW ON: 75% Qr = 510kVAr OFF: 25% Qr = 170kVAr Threshold Controller Margin Controller Reserve Capacity = (α × load kW) + (β × wind kW) Controller turns on another diesel if present generator cannot meet the current load plus the Reserve Capacity. Controller turns diesel off if it is not needed to meet the current load plus the Reserve Capacity.

11. www.risoe.dk Hybrid System Performance Evaluation Diesel Generator Fuel Curve No-load offset

12. www.risoe.dk Hybrid System Performance Evaluation Graphical Output from IPSYS

13. www.risoe.dk Hybrid System Performance Evaluation Running of the Diesel Generators 1 min 10 min 60 min Active Power (kW)

14. www.risoe.dk Hybrid System Performance Evaluation Numbers of Starts for Diesel No. 2 and Hours Run

15. www.risoe.dk Hybrid System Performance Evaluation IPSYS Outputs for this Assessment Total energy generated during the simulation Wind energy generated Wind usage Wind share Dump load share Diesel generator starts Diesel generator run time Total diesel fuel used

16. www.risoe.dk Hybrid System Performance Evaluation Simulation Cases Investigated Base Case Threshold Controller & Raw Wind Input Threshold Controller with min runtime adjustments & Raw Wind Input Threshold Controller & Filtered Wind Input All simulations run with 1, 10 & 60 minute timesteps Margin Controller & Raw Wind Input

17. www.risoe.dk Hybrid System Performance Evaluation Accuracy is necessary in modelling limits in even simple controllers. Both controllers show a sensitivity to a change in limits

18. www.risoe.dk Hybrid System Performance Evaluation ….but the timestep can influence the effect. The effect of changing the minimum run time is reduced as the timestep is increased.

19. www.risoe.dk Hybrid System Performance Evaluation Timestep impact on long-term performance indicators is relatively small

20. www.risoe.dk Hybrid System Performance Evaluation Major impact of timestep is on diesel start/stops

21. www.risoe.dk Hybrid System Performance Evaluation Impact of Wind Sampling Period (1) Raw Wind Data Series Power Frequency Filter

22. www.risoe.dk Hybrid System Performance Evaluation Impact of Wind Sampling Period (2)

23. www.risoe.dk Hybrid System Performance Evaluation Conclusions Simulation timestep has only a limited impact on overall system performance in terms of fuel consumption and wind usage Simulation timestep has a significant impact on operating conditions of components e.g. number of start/stops of genset Wind filtering has only minimal impact on fuel consumption For feasibility studies long timesteps (1h) give results of adequate accuracy Short timesteps (1min) are necessary for study of controllers and interaction between components, component operating conditions and wear and tear It is important to model reactive power aspect of controllers, as this has a considerable impact on the system operation and simulation results. Short time steps will allow real controllers to be implemented in simulation tool – IPSYS is designed for that

24. www.risoe.dk Hybrid System Performance Evaluation Limitations on Results Only one hybrid system configuration has been simulated. No storage included in studied configuration. Only simple controllers investigated. System parameters were adjusted to ensure all runs at all timesteps would run – this could hide the full of impact of timestep changes. Only a short step towards system testing – Lots of work required