Programa de Atração de Jovens Talentos - Nível A: Robust and probabilistic models for electricity power systems operation and planning with significant renewable energy integration

Programa de Atração de Jovens Talentos - Nível A - 2014

Bolsista: David Pozo

Coordenador: Alexandre Street

Entidade financiadora: CAPES-CNPq-MCTI

SUMMARY

Decision-making under uncertainty plays a key role in the operation and planning of restructured power systems. Traditional power systems have been modeled with optimization problems to account for very little uncertainty sources as demand or Renewable Energy Sources (RES). Then, the conventional two-stage stochastic optimization problems have been enough to operate and plan power systems. However, increasing share of RES in the generation mix of power systems presents a big challenge in the efficient management, mainly due to the limited predictability and the high variability of renewable generation, features that make RES plants non-dispatchable. The evolution of electricity systems indicates an increasing integration of RES. This means more production volatility, forcing conventional generators to modify their productions according to the intermittence of the RES. A few economies, which have been recognized as leadership in incentives for RES, are now revising their RES target levels due to huge integration costs incurred in the last decade. Network reinforcements, due to deviations from optimal planning decisions, and preventive as well as corrective operative actions, due to the high level of reserves and redispatch of fast but expensive units needed to cope with RES intermittence are major causes for this tendency. As a consequence, probabilistic and robust optimization models constitute an alternative to cope with future renewable energy penetration exploring theirs synergies while ensuring required well-known power system standards. Thus, the authors of this project believe that the further development of operative and planning models to account for such variability can lead to a more reliable and secure operation of the power system in electricity grids with significant RES penetration.

Typically, the following guidelines are desired in power systems operation and planning decisions: (a) economic efficiency, so as to minimize the cost of providing power; (b) supply reliability criteria, where the energy reaches all clients with minimal service interruptions; (c) freedom of agents and transparency of processes and markets (mainly in liberalized markets); (d) increasing renewable energy sources; (e) optimal investment incentive based economic models resulting from liberalized markets (such as the spot market).

As argued before, high level of RES in power systems, which is in line with guideline (d), may jeopardizes power systems supply reliability and their economic efficiency, guidelines (b) and (a), respectively. Therefore, the main objective in this project is to design, develop and analyze alternative robust and probabilistic models in operational (unit commitment) and planning (transmission expansion) framework, to incorporate the effect of RES under current security standards. To do that, the incorporation of tighter security criteria, such as the n-K criterion under the full network representation, as well as the renewable energy sources temporal and spatial uncertainty correlation is needed. This problem affects most of the world's electricity markets, including Brazil. The following novel features characterize these alternative procedures:

  1. Robust optimization models of unit commitment and transmission expansion planning (TEP) problems. This aspect is particularly relevant to properly consider scenario-free uncertainty sets. Likewise, these novel models include relevant information on the uncertainty set related with temporal and spatial correlation.
  2. Probabilistic optimization models of unit commitment and TEP problem. Probabilistic optimization integrates reliability constraints to obtain effective solutions for highly variable uncertainty sets.
  3. Extension of the traditional n-1 and n-2 security criteria. The new n-K security criterion considers up to K simultaneous contingencies.
  4. Consideration of renewable energy with high levels of integration.
  5. Development of decomposition techniques to improve the tractability of the proposed models.

The proposed research effort will result in both better mathematical models and improvements in the required solution techniques. The development of these tools constitutes challenging research work. Moreover, the international power engineering scientific community is particularly interested in this research, as can be seen in major technical journals and conferences. Additionally, expected results are of high interest for the electric power industry. Finally, it is important to say that this subject has come to the community’s attention very recently, in 2010, with the first paper published in the IEEE TPWRS (Street et. al 2010) [7]. After that, [8], [9], and [14] extended the findings of [7] to account for relevant features of power systems operation. The aforementioned publications demonstrate the correlation between the candidate and the technical coordinator research themes, corroborating, thus, a promising gain of synergy.