Tractable Stochastic Unit Commitment for Large Systems During Predictable Hazards

Farshad Mohammadi,Mostafa Sahraei-Ardakani

doi:10.1109/access.2020.3004391

Farshad Mohammadi, Mostafa Sahraei-Ardakani

Open Access

https://doi.org/10.1109/access.2020.3004391

Copy DOI

Abstract

Many foreseeable natural hazards, including extreme weather events, lead to an outage of multiple transmission lines. Although such outages can be predicted in advance, there is a great deal of uncertainty in these predictions. To appropriately use the failure estimations in power system scheduling, this paper formulates a stochastic unit commitment (SUC) problem with explicit modeling of the predicted outages. The formulated problem, however, is extremely computationally-demanding, as the uncertainty is placed on the binary status of transmission lines. This paper, then, develops a computationally efficient algorithm to solve the formulated SUC for large-scale systems. The algorithm employs generation shift factors to enable rapid calculation of power flows. Additionally, flow canceling transactions are used to model multiple line outages without having to recalculate shift factors. Finally, critical constraints are iteratively detected and added to the problem. This approach substantially reduces the size of the problem, which helps computational tractability. The effectiveness of the developed algorithm is demonstrated through simulation studies on the Texas 2000-bus test system. The algorithm is used to minimize the lost load during a hypothetical hurricane. The results show that the algorithm is computationally tractable and can effectively identify a preventive dispatch, leading to a substantial reduction in power outages.

Highlights

Short term reliability and economic efficiency in power systems are achieved through scheduling of the available resources
This paper offers a solution to this challenge by developing a computationally tractable model for solving the unit commitment problem, in the presence of predictable but uncertain possibility of multiple line outages
The main contribution of this paper is developing a computationally efficient stochastic unit commitment formulation, which can handle multiple uncertain but predictable line outages. This is achieved through the adoption of flow canceling transactions [30], which enables fast calculation of power flows even in the presence of multiple line outages

Summary

INTRODUCTION

Short term reliability and economic efficiency in power systems are achieved through scheduling of the available resources. This paper offers a solution to this challenge by developing a computationally tractable model for solving the unit commitment problem, in the presence of predictable but uncertain possibility of multiple line outages. As stochastic optimization explicitly models uncertainties, it can achieve higher levels of reliability and economic efficiency This improvement in solution quality, comes at the cost of substantial computational burden [1], [9]–[11]. The main contribution of this paper is developing a computationally efficient stochastic unit commitment formulation, which can handle multiple uncertain but predictable line outages. This is achieved through the adoption of flow canceling transactions [30], which enables fast calculation of power flows even in the presence of multiple line outages.

MODELING TECHNIQUES

MULTIPLE LINE OUTAGE MODELING

Objective

PROBABILISTIC MODEL OF LINE OUTAGES

Findings

CONCLUSION