Optimal power flow problem considering multiple-fuel options and disjoint operating zones: A solver-friendly MINLP model

Author(s)

Pourakbari-Kasmaei, Mahdi ^[1] ; Lehtonen, Matti ^[1] ; Fotuhi-Firuzabad, Mahmud ^{[1, 2]} ; Marzband, Mousa ^{[3, 4]} ; Sanches Mantovani, Jose Roberto ^[5]

Affiliation

^[1] Aalto Univ, Dept Elect Engn & Automat, Maarintie 8, Espoo 02150 - Finland

^[2] Sharif Univ Technol, Dept Elect Engn, Tehran - Iran

^[3] Northumbria Univ Newcastle, Dept Maths Phys & Elect Engn, Fac Engn & Environm, Newcastle Upon Tyne, Tyne & Wear - England

^[4] Islamic Azad Univ, Lahijan Branch, Dept Elect Engn, Lahijan - Iran

^[5] State Univ Sao Paulo UNESP, Dept Elect Engn, Ilha Solteira - Brazil

• Document type: Review article
• Source: INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS; v. 113, p. 45-55, DEC 2019.
• Web of Science Citations: 5

Abstract

This paper proposes a solver-friendly model for disjoint, non-smooth, and nonconvex optimal power flow (OPF) problems. The conventional OPF problem is considered as a nonconvex and highly nonlinear problem for which finding a high-quality solution is a big challenge. However, considering practical logic-based constraints, namely multiple-fuel options (MFOs) and prohibited operating zones (POZs), jointly with the non-smooth terms such as valve point effect (VPE) results in even more difficulties in finding a near-optimal solution. In complex problems, the nonlinearity itself is not a big issue in finding the optimal solution, but the nonconvexity does matter and considering MFO, POZ, and VPE increase the degree of nonconvexity exponentially. Another primary concern in practice is related to the limitations of the existing commercial solvers in handling the original logic-based models. These solvers either fail or show intractability in solving the equivalent mixed integer nonlinear programming (MINLP) models. This paper aims at addressing the existing gaps in the literature, mainly handling the MFOs and POZs simultaneously in OPF problems by proposing a solver-friendly MINLP (SF-MINLP) model. In this regard, due to the actions that are done in the pre-solve step of the existing commercial MINLP solvers, the most adaptable model is obtained by melting the primary integer decision variables, associated with the feasible region, into the objective function. For the verification and didactical purposes, the proposed SF-MINLP model is applied to the IEEE 30-bus system under two different loading conditions, namely normal and increased, and details are provided. The model is also tested on the IEEE 118-bus system to reveal its effectiveness and applicability in larger-scale systems. Results show the effectiveness and tractability of the model in finding a high-quality solution with high computational efficiency. (AU)

• FAPESP's process: 16/14319-7 - Online carbon footprint Allocation- a joint Price- and carbon Footprint: responsive demands model to reduce the GHG emissions and manage the electricity demand
• Grantee: Mahdi Pourakbari Kasmaei
• Support Opportunities: Scholarships abroad - Research Internship - Post-doctor
• FAPESP's process: 15/21972-6 - Optimization of the operation and planning in transmission and distribution systems
• Grantee: Rubén Augusto Romero Lázaro
• Support Opportunities: Research Projects - Thematic Grants
• FAPESP's process: 14/22828-3 - A mixed-integer nonlinear programming paradigm to solve multi fuel-based environmentally-constrained active-reactive optimal power flow
• Grantee: Mahdi Pourakbari Kasmaei
• Support Opportunities: Scholarships in Brazil - Post-Doctoral