A Modified DistFlow for Distributed Generation Planning Problems in Radial Grids - Archive ouverte HAL
Communication Dans Un Congrès Année : 2020

A Modified DistFlow for Distributed Generation Planning Problems in Radial Grids

Résumé

This paper focuses on a simplified expression of the load flow constraints for the optimal planning of distributed generation (DG) in radial grids. The proposed approach lies on the conventional DistFlow equations improved with the consideration of linearized bi-directional losses and their impact on the voltage profiles. The method is validated on various radial test cases and allows to run time coupled power flows in order to consider dynamic constraints (e.g. battery state of chare) in short computational times. With the help of the proposed method, a DG placement problem is solved in less than 2 minutes with an exhaustive search over a 69 buses system and the operation is simulated along a single representative day. A scalability test is finally performed with longer time horizons and greater numbers of connected DG. The developed approach allows to run the operation of a 69 buses system over 2 months and with 20 connected DGs in less than 5 minutes. I. NOMENCLATURE Sets : t ∈ T set of time steps b ∈ B set of buses (B buses) l ∈ L set of branches (L lines) k ∈ K set of piece wise segments (K segments) l d(b) ∈ D(b) set of lines downstream a bus b l u(b) upstream line of bus b (one element) b d(l) downstream/end bus of line l (one element) b u(l) upstream/start bus of line l (one element) Variables (in p.u.): For l=l d(b) l u(b) p l,t ,q l,t active and reactive powers in line l at time t δp l,t ,δq l,t active and reactive powers in line l at time t V b,tline l reactive losses (+/-) in line l at time t P b,t , Q b,t active and reactive injected at bus b at time t P b,t bat+ , P b,t bat-charge/discharge of battery at bus b at time t SOC b,t bat state of charge of battery at bus b at time t (%) Parameters : P b,t ,Q b,t active and reactive load at bus b at time t p ̅ l , q ̅ l line l active reactive limit power r l ,x l resistance and reactance of line l E bat , bat battery capacity at bus b P b,t solar generation at bus b at time t II. INTRODUCTION The benefits of distributed generation (DG) integrated in conventional power systems have long been identified-with losses and emissions reduction, improved voltage profiles and enhanced system reliability as well as the opportunity to defer significant grid investments [1]. Oftentimes, in the framework of DG planning studies, the objective is to find the best type, location and size of the considered resources that may be renewable based or fossil fueled generators and energy storage units [2]. One of the main technical challenges encountered is to avoid prohibitive computational times. Indeed, when integrated design and management problems are investigated, the operation of the systems has to be simulated for a great number of configurations before finding the optimal DG plan [3]. Especially, when storage devices are considered, the system operation is estimated over a representative period (day, month, year). With energy constraints linked across the successive time steps, the power flow equations have to be simulated over the whole horizon, and typical AC power flow (ACPF) cannot be run successively for every time step, as seen in [4]. One way to leverage that complexity is to consider DC power flow (DCPF) constraints [5] at the cost of extensive simplifications that are not relevant for distribution networks, to which most of the small distributed resources are expected to be connected. In particular, reactive power and voltage drop are not considered. For radial systems, a simplified DistFlow approach allows to efficiently compute the active/reactive power flows while updating the voltage drop form the slack bus (typically the point of common coupling, PCC) to the termination node of every feeder [6]. Linearized approaches of the DistFlow are encountered in planning studies or in the presence of distributed resources for faster computation [7] [8]. Similarly to the DCPF equations, those approaches lie on strong assumptions, the line losses being oftentimes neglected when computing the branch flow or bus voltages. This paper proposes a modified version of the linearized DistFlow to overcome those shortcomings. A particular attention is attached to the consideration of bi-directional active/reactive losses in the context of resources allocation that may incur backfeed power. The method is then validated over various test cases before a generic, simple DG planning problem is considered. The main contribution of the work are:  A DistFlow formulation that considers linearized bi-directional losses and their impact on the voltage.  The validation of the method for a simple allocation problem solved in less than 2 min in a 69 bus system.  A scalability test for longer planning horizons and greater numbers of distributed resources.
Fichier principal
Vignette du fichier
Paper - Modified DistFlow_vF2.pdf (1.02 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03016130 , version 1 (20-11-2020)

Identifiants

Citer

Remy Rigo-Mariani, Vincent Debusschere, Marie-Cécile Alvarez-Hérault. A Modified DistFlow for Distributed Generation Planning Problems in Radial Grids. IECON 2020 - 46th Annual Conference of the IEEE Industrial Electronics Society, Oct 2020, Singapore, Singapore. pp.1626-1632, ⟨10.1109/IECON43393.2020.9254865⟩. ⟨hal-03016130⟩
72 Consultations
700 Téléchargements

Altmetric

Partager

More