1

Midterm ReportReporter : Yu

2

Operational Adequacy Studies of Power Systems With Wind Farms and Energy Storages

Base-Load Cycling on a System With Significant Wind Penetration

Outline

3

Operational Adequacy Studies of Power Systems With Wind Farms and Energy Storages

Peng Wang,Member, IEEE, Zhiyong Gao, Student Member, IEEE, and Lina Bertling Tjernberg, SeniorMember, IEEE

First paper

4

The growth rate of global installed wind power capacity was about 30% per annum in the last decade

High renewable power penetration has significant impacts on system stability and reliability due to the intermittent and uncertain characteristics of wind speed and solar irradiation.

Nowaday

5

SR is defined as the operating reserve provided by CGs that are synchronized with the power system and can pick up load immediately during the failures of any committed generators

Shortage of SR may reduce system reliability. On the other hand, over committed spinning reserve will cause energy waste.

Temporarily Solution

6

The simulation results show that the ramp rates of CGs have large impacts on both EENU and EENS. The incorporation of wind farms without ESS has very limited system reliability improvement while leads to large amount of energy surplus. ESS can significantly improve the wind power capacity credit of a wind farm. However there is maximum size of ESS for a given size of a wind farm and a given wind speed model. Reliability and cost should be balanced when selecting the ESS.

Conclusion

7

The autoregressive moving average (ARMA) time-series model is usually adopted to predict wind speed

Most wind speed data obtained from practical wind farms andweather stations are averagewind speed for a given period such as 5, 10, 30 or 60 min as shown in Fig. 1.

Wind Speed Model

8

WS is real wind spewed. Wsci and Wsco are the cut-in and cut-out

wind speed WSr is the rated wind speed Pr is the rated power of the WTG A,B, and C are WTG parameters

Power Output Model of a WTG

9

In system operation, a time varying load model as shown in Fig. 2 is usually used to simulate real time characteristics of load.

Load Model

10

Battery storage can be controlled to charge or discharge in constant or variable rate based on system operation requirements. The advantage of battery ESS is its fast charging/discharging rate.

Energy Storage Model

12

The basic assumption for this model is that CGs can be committed or shut down immediately when required.

The ramp rates of CGs may cause energy deficiency or surplus. For a large power system with low wind power penetration, the effect of ramp rates can be ignored.

Ramp Rates of a Generator

20

Step 1) Input wind speed, load level, reliability data for WTGs and CGs, and initial SOC of ESS.

Step 2) For the operating period , determine committed CGs based on unit loading priority order.

Step 3) For system contingency state , calculate state probability.

Step 4) Calculate and using the developed formulas based on the load and generation condition, and update the total EENS and EENU.

Step 5) If all the states are considered, go to Step 6; otherwise go to Step 3.

Step 6) If all the periods are considered, go to Step 7; otherwise go to step 2.

Step 7) Output the results.

Evaluation Procedure

21

The total installed capacity of CGs is 3405MWand the annual peak load is 2850 MW. In order to illustrate the impact ofWTGs and ESSs, the peak load of the RTS is increased to 3135 MW. The lead time is assumed to be 1 h.

SYSTEM STUDIES

22

Case 1: In this case, the ramp rates of CGs are not considered, i.e., the ramp up/down is modeled as a step function. Priority order I is used for unit commitment. The economic generation dispatch is based on the operating costs of the units. The spinning reserve is 15% of the peak load. The monthly EENS and EENU in this case are 0.25 MWh and 0 MWh, respectively. There is no energy surplus when the CG has no ramp rate constraints.

System With Only CGs

23

Case 2: The ramp rates of 10% of committed generation capacity are considered in this case. The EENS and EENU depend on the number and types of the power output of the committed units. When the ramp capacity is provided only by spinning units, the monthly EENS is 126.32 MWh and the EENU is 7981.68 MWh because of the low system ramp rates. If the system ramp capacity can be provided by all committed units in a period, the EENS and EENU are reduced to 6.25 MWh and 1356.89 MWh, respectively.

System With Only CGs

24

Basic Case: Awind farm with 100 identical 5-MW WTGs is added into the system in Section VI-A. The unreliability U of a WTG is assumed to be 0.01. The cut-in, rated and cut-out wind speeds are 3 m/s, 14 m/s, and 25 m/s, respectively. The total wind energy output of the wind farm in June is 123000 MWh. With consideration of ramp rates of CGs, the EENS is reduced 43%from6.25MWhwithout WTGs to 3.59MWh and the EENU is increased 44% from 1356.89 MWh to 2054.34 MWh. The reliability of the system is improved. However, the energy surplus is increased at the same time with the addition

System With Both CGs and WTGs

25

The impact of wind power penetration level on system EENS and EENU is simulated in this case. When wind power penetration level increases from 5% to 24% of the total CG capacity to supply the same load in Section VI-A, the curves of EENU and EENS are shown in Fig. 9. Although EENS is reduced from 4.71 MWh to 3.12 MWh, EENU is significantly increased from 1415.32 MWh to 2913.88 MWh. The results indicate that less reliability improvement leads to more wind energy being wasted, which means it is not worth (economical) to add more WTGs to improve reliability.

Effect of Wind Power Penetration Level

26

Result

27

When the SR increases from 5% to 24% based on the system operating condition in Section VI-A, the corresponding curves of EENU and EENS are shown in Fig. 10. It is obvious that the EENS decreases with the increase of SR. When SR is over 15%, the reduction of EENS becomes smaller for same percentage increase of SR. The EENS is almost constant when SR reaches 19%. However the EENU increases almost linearly with the increase of SR, which means that it is not economic to continuously increase the SR to improve system reliability.

Effect of SR

28

Result

29

To replace two 50-MW CGs for the system, 152* 5 MW WTGs are required to achieve the same EENS of 6.25 MWh/month. It is obvious that a WTG cannot provide the same capacity credit as a same size CG. On the other hand, the EENU increasea significantly from 1356.89 MWh to 1789.56 MWh.

Rate ESS Capacity

30

With Particular Wind Speed

31

If the wind farm has the higher average wind speed of 14 m/s and less intermittency as shown in Fig. 11, 40*5 MW WTG units is required to replace two 50-MW CGs for the same EENS. The EENU is reduced from 1789.56 MWh to 1701.28 MWh.

With Particular Wind Speed

32

Case 1: A 100-MWh battery ESS is added into the system based on the system condition in Section VI-B1. The charging/ discharging rate is 100 MW (100 MWh/hour). The monthly EENS is reduced 86% from 3.59 MWh without the ESS to 0.41 MWh with the ESS. The monthly EENU is reduced 16% from 2054.34MWh to 1723.91MWh. The EES also reduces the system EENU due to its fast ramp (charging/discharging) rates compared with CGs.

System With CGs, WTGs, and ESS

33

Case 2: For the fixed discharging rate 100 MW, the effects of the ESS capacity on the EENS and EENU are investigated in this case. When the EES capacity changes from 10MWh to 200 MWh, the EENU and EENS are shown in Fig. 12. It can be seen from Fig. 12 that both the EENS and

System With CGs, WTGs, and ESS

34

Result

35

Case 3: Fig. 13 shows that the required WTG capacity to replace two 50-MW CGs for different ESS capacities under a given EENS.

System With CGs, WTGs, and ESS

36

Result

37

Base-Load Cycling on a System With Significant Wind Penetration

Niamh Troy, Graduate Student Member, IEEE, Eleanor Denny, Member, IEEE, and Mark O’Malley, Fellow, IEEE

Second Paper

38

increasing penetration of variable power sources coupled with the deregulation of the electricity sector could lead to these base-load units being shut down or operated at partload levels more often.

This cycling operation would have onerous effects on the components of these units and potentially lead to increased outages and significant costs.

This paper shows the serious impact increasing levels of wind power will have on the operation of base-load units.

Finally, it is shown that if the total cycling costs of the individual base-load units are taken into consideration in the scheduling model, subsequent cycling operation can be reduced.

Abstract

39

even when state-of-the-art methods of forecasting are employed, the next day hourly predicted wind output can vary by 10%–15% of the total wind capacity.

in certain systems wind is allowed to self-dispatch, so forecast output is not included in the day-ahead schedule. This can lead to increased transmission constraints which will further intensify plant cycling and has been shown to displace energy from combined cycle gas turbines (CCGTs) in particular.

Introduction

40

As these units are designed with minimal operational flexibility, cycling these units will result in accelerated

deterioration of the units’ components through various degeneration mechanisms such as fatigue, erosion, corrosion, etc, leading to more frequent forced outages and loss of income. The start/stop operation and varying load levels result in thermal transients being set up in thick-walled components placing them under stress and causing them to crack. Hence cycling of base-load units will impose additional costs on the unit, the most apparent being increased operations and maintenance (O&M) and capital costs resulting from deterioration of the components. However, fuel costs will also increase with cycling operation as the unit will be starting up more frequently, and also because the overall efficiency of the unit will deteriorate

Nowaday

41

Simulations were carried out using a scheduling model called the Wilmar Planning Tool.

It combined the benefits of mixed integer optimization with stochastic modeling. The main functionality of the Wilmar Planning Tool is embedded in the Scenario Tree Tool and the Scheduling Model.

Model Program

42

The Scenario Tree Tool generates scenario trees containing three inputs to the scheduling model: wind, load and demand for replacement reserve. Realistic possible wind forecast errors are generated using an auto regressive moving average (ARMA) approach which considers the historical statistical behavior of wind at individual sites. Historical wind speed series taken from the various sites are then added to the wind speed forecast error scenarios to generate wind speed forecast scenarios. These are then transformed to wind power forecast scenarios. Load forecast scenarios are generated in a similar manner.

Model Program

43

reserve is categorized as primary or replacement. Primary reserve, which is needed in short time scales (less than five minutes), is supplied only by synchronized units. The system should have enough primary reserve to cover an outage of the largest online unit occurring at the same time as a fast decrease in wind power production. A forced outage time series for each unit is also generated by the scenario tree tool using a semi-Markov process based on given data of forced outage rates, mean time to repair and scheduled outages is produced. Any unit that is offline and can come online in under one hour can provide replacement reserve.

Model control

44

The 2020 Irish system was chosen as a test case for this study because its unique features make it suitable for investigating base-load cycling. It is a small island system, with limited interconnection to Great Britain, a large portion of base-load plant and significant wind penetration. Thus, potential issues with cycling of base-load units may arise on this system at a lower wind penetration.

Model control

45

The test system includes four 73 MW pumped storage units with a round-trip efficiency of 75% and a maximum pumping capacity of 70MW each and two 83 MW CHP units with “must-run” status as they provide heat for industrial purposes.

System Parameter

46

carbon price of 30/ton was assumed

System Parameter

47

the model was run stochastically, for one year, for the “no wind” case and each of the three wind cases, first, without any pumped storage on the system and second, without any interconnection on the system. In order to fairly compare systems without storage/interconnection to the systems with storage/interconnection

System Set

48

Number of thermal units online with increasing wind penetration(averaged at each hour shown over a two-week period in april)

System Set

49

Moving from 0% to 42% installed wind capacity the annual start-ups for a typical CCGT unit rise from 22 to 98, an increase of 340%. This increase in CCGT start-ups corresponds to a plummeting capacity factor as seen in Fig. 1. Thus increasing levels of wind effectively displaces CCGT units into mid-merit operation

FIG 1

50

Result

51

Result

52

Result

53

Result

54

Result

55

Result

56

Result

59

Thanks for Listening