Reserve procurement in power systems with high renewable capacity: - - PowerPoint PPT Presentation

Outline The EU context Model description Case Study Results Results

Reserve procurement in power systems with high renewable capacity: How does the time framework matter?

• G. Oggioni(1)
• R. Dominguez(2)
• Y. Smeers(3)

(1) University of Brescia, Italy (2) Universidad de Castilla-La Mancha, Toledo, Spain (3)CORE, Universit´ e catholique de Louvain, Belgium

Mercati energetici e metodi quantitativi: un ponte tra Universit` a ed Impresa Padova October 13th, 2016

Reserve procurement

Outline The EU context Model description Case Study Results Results

Outline

1 The European context 2 Model description

Model common assumptions Model 1: joint procurement of energy and reaserve Model 2: reserve procured before day ahead Model 3: reserve procured after day ahead

3 Case Study 4 Results 5 Conclusions Reserve procurement

Outline The EU context Model description Case Study Results Results

Reserve procurement and RES integration

Renewable energy integration requires flexibility because of: Uncertainty; Variability. The schedule of an adequate reserve level is becoming extremely important because: The increasing integration of stochastic (renewable) energy production makes power systems unstable It guarantees security of supply and system balance in real time!

Reserve procurement

Outline The EU context Model description Case Study Results Results

Towards the Internal European Electricity Market

Third Energy Package and Network Codes

The European Commission envisages the coordination of:

The energy day-ahead markets (Price Coupling of Regions); The reserve procurement mechanisms; The congestion management; The energy balancing markets.

Reserve procurement

Outline The EU context Model description Case Study Results Results

Goals

Reserve procurement

Outline The EU context Model description Case Study Results Results

Our goals in this paper...

Q1: Does the time framework for reserve procurement matter? We analyze and compare the efficiency levels of three power systems where:

1 Energy and reserves are jointly scheduled by an Independent System Operator

(as in the US)

2 Reserves are scheduled before the clearing of the day-ahead energy market

(as in Central European countries)

3 Reserves are schedule after the clearing of the day-ahead energy market

(as in Italy, Spain, Portugal)

Q2: Does a coordinated reserve procurement increase the system efficiency? We compare the efficiency levels of the three power systems above assuming a coordinated and not-coordinated reserve schedule.

Reserve procurement

Outline The EU context Model description Case Study Results Results

Models

Reserve procurement

Outline The EU context Model description Case Study Results Results Common assumptions

Model common assumptions

Spatial granularity: nodal level both in day-ahead energy and ancillary service markets Reserves: Conventional and downward/upward spinning reserves Generating units: Stochastic (wind and solar PV) vs. dispatchable units (nuclear, coal, CCGT)

Dispatchable units Qualified Non-qualified Coal Nuclear CCGT

Demand response: demand side management with downward/upward deviations in real time Uncertainty characterization: day-ahead forecasts and real time scenarios for demand level and renewable power availability

Reserve procurement

Outline The EU context Model description Case Study Results Results Model 1

Model 1: Energy and reserve needs are jointly scheduled

Model 1 is a two-stage stochastic programming problem as illustrated below:

First Stage

Second Stage D-1 (Day ahead)

ISO balances the system on the basis of RT scenarios ISO co-optimizes the energy and the reserve procurement

s1 s2 s3 D (Real time)

Figure: Decision-making process of Model 1

Reserve procurement

Outline The EU context Model description Case Study Results Results Model 2

Model 2: Reserve scheduled before the day-ahead energy market

Model 2 is a three-stage stochastic programming problem as illustrated below:

First Stage Second Stage Third Stage

f1

PX clears the energy market TSO procures reserves PX clears the energy market PX clears the energy market TSO balances the system on the basis

• f RT scenarios

f2 f3 S1f1 S2f1 S3f1 S1f2 S2f2 S3f2 S1f3 S2f3 S3f3

W-1 (Week ahead) D-1 (Day ahead) D (Real time)

Figure: Decision-making process of Model 2

Reserve procurement

Outline The EU context Model description Case Study Results Results Model 3

Model 3: Reserve scheduled after the day-ahead energy market

Model 3 is formulated as illustrated below:

First Stage Second Stage TSO balances the system on the basis of RT scenarios TSO re-dispatches energy and procures reserves PX clears the energy market s1 s2 s3 D-1 (Day ahead) D (Real time) D-1 (Day ahead)

Figure: Decision-making process of Model 3

Reserve procurement

Outline The EU context Model description Case Study Results Results

Case Study

Reserve procurement

Outline The EU context Model description Case Study Results Results

Case study

Nodal network: IEEE 24-node network with 38 transmission lines Capacity:

Technology Capacity (MW) CCGT 2250 Coal 700 Nuclear 900 Wind 2100 Solar 750 Total 6700

Total demand (17 nodes): 3135 MW Uncertainty: 3 day-ahead forecasts and 3 real time scenarios per day-ahead forecast

18 21 22 17 16 19 20 23 15 14 13 11 12 24 3 9 10 6 4 5 2 1 7 8 W W CCGT CCGT N PV W CCGT CCGT PV W CCGT PV W W CCGT CO

Z2 Z1 Z3

PV CO

Reserve procurement

Outline The EU context Model description Case Study Results Results

Reserve procurement

Coordinated procurement: Reserve need is determined on the whole market as a unique zone (1 zone); Not-coordinated procurement: Reserve needs are defined at zonal level (3 zones/countries).

Reserve procurement

Outline The EU context Model description Case Study Results Results

Results

Reserve procurement

Outline The EU context Model description Case Study Results Results

Operating costs

Coordinated reserve procurement ($):

1 Zone Model 1 Model 2 Model 3 (Expected) (Expected) Total operating costs 826,180 837,708 827,296 DA operating costs 822,345 851,960 823,532 RT operating costs 3,835

• 14,252

3,764

Not-coordinated reserve procurement ($):

3 Zones Model 1 Model 2 Model 3 (Expected) (Expected) Total operating costs 834,007 843,395 5,060,361 DA operating costs 829,937 860,962 858,323 DA unserved demand value

• 4,201,153

RT operating costs 4,070

• 17,568

884 RT unserved demand value

• Not-coordinated reserve procurement and increased installed capacity ($):

3 Zones (Increased capacity) Model 1 Model 2 Model 3 (Expected) (Expected) Total operating costs 772,771 775,675 776,195 DA operating costs 786,711 798,431 792,769 RT operating costs

• 13,940
• 22,756
• 16,574

Reserve procurement

Outline The EU context Model description Case Study Results Results

Conclusions

Reserve procurement

Outline The EU context Model description Case Study Results Results

Conclusions

As expected, the market structure represented through Model 1 (one ISO) results as the most efficient market under all reserve procurement assumptions. We also verified that the not-coordinated reserve procurement based on multiple reliability zones leads to higher total operating costs than considering the power system as a whole. Model 3 in the coordinated reserve procurement case results almost as efficient as Model 1. But it becomes inefficient (unserved demand) in the not-coordinated reserve procurement because of the limits imposed on the cross-border exchanges.

Reserve procurement

Outline The EU context Model description Case Study Results Results Reserve procurement

Outline The EU context Model description Case Study Results Results

Morales, J.M., Conejo, A.J., Madsen, H., Pinson, P., Zugno, M. (2014). Integrating renewables in electricity markets: Operational Problems. International series in

• perations research and management science: 205. New York, NY, USA: Springer.

Fabbri, A., Gomez San Roman, T., Rivier Abbad, J., Mendez Quezada, V. H. (2005). Assessment of the cost associated with wind generation prediction errors in a liberalized electricity market, IEEE Transaction on Power Systems, 20(3), 1440-1446. Ortega-Vazquez, M.A., Kirschen, D.S. (2009). Estimating the Spinning Reserve Requirements in Systems With Significant Wind Power Generation Penetration, IEEE Transaction on Power Systems, 24(1), 114-124. Papavasiliou, A., Oren, S.S., O’Neill, R.P. (2011). Reserve requirements for wind power integration: a scenario-based stochastic programming framework. IEEE Transaction on Power Systems, 26(4), 2197-2206. Pineda, S., Morales, J.M. (2016). Capacity expansion of stochastic power generation under two-stage electricity markets, Computers and Operations Research, 70, 101-114. Reliability Test System Task Force (1999). The IEEE reliability test system-1996, IEEE Transaction on Power Systems, 14(3), 1010-1020.

Reserve procurement