Enhanced Flexibility Aggregation Using LinDistFlow Model with Loss Compensation

Yanlin Jiang 2 , Xinliang Dai 2 , Frederik Zahn 2 and Veit Hagenmeyer 2 · 2025

Read Enhanced Flexibility Aggregation Using LinDistFlow Model with Loss Compensation by Yanlin Jiang 2 , Xinliang Dai 2 , Frederik Zahn 2 and Veit Hagenmeyer 2 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-01715.txt.

Introduction

Grid operators who link home solar and batteries to the main grid must share plans without sharing private data. Yanlin Jiang and three co-authors work on that join in 2025, through a method that sums what a local grid can flex. The sum rests on a fast, straight-line stand-in for true power flow, which bends the truth in known ways.

The team lays out the stand-in's errors and shows how they warp the flex sum. Their fix adds a payback term that trims those errors in a simple, direct way. Tests show the refined sum lets the two grid levels work as one with less strain. As green power spreads, such sums are a key tool for grids that must bend without breaking.

Catalog record

CREATORYanlin Jiang 2 , Xinliang Dai 2 , Frederik Zahn 2 and Veit Hagenmeyer 2
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE28 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-01715

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Enhanced Flexibility Aggregation Using LinDistFlow Model with Loss Compensation Thanks: Corresponding: Xinliang Dai Yanlin Jiang 2 , Xinliang Dai 2 , Frederik Zahn 2 and Veit Hagenmeyer 2 Abstract With the increasing integration of renewable energy resources and the growing need for data privacy between system operators, flexibility aggregation methods have emerged as a promising solution to coordinate integrated transmission-distribution (ITD) systems with limited information exchange. However, existing methods face significant challenges due to the nonlinearity of AC power flow models, and therefore mostly rely on linearized models. This paper examines the inherent errors in the LinDistFlow model, a linearized approximation, and demonstrates their impact on flexibility aggregation. To address these issues, we propose an intuitive compensation approach to refine the LinDistFlow-based flexibility set. Simulation results demonstrate the effectiveness of the proposed method in efficiently coordinating ITD systems. I Introduction In recent years, the share of distributed energy resources (DERs) in global power generation has increased steadily. This transition poses significant challenges to traditional centralized coordination frameworks, primarily due to the growing complexity of system models introduced by the addition of DERs , raising privacy concerns and increased demands on communication [ 1 , 2 , 3 ] . To address these challenges, significant efforts have focused on developing solutions to efficiently coordinate modern integrated transmission-distribution (ITD) systems. One promising approach is power flexibility aggregation, which leverages the collective flexibility of DERs to facilitate transmission-distribution interaction [ 4 , 5 , 6 ] . Flexibility refers to the set of all feasible power flows at the interface between transmission system operator (TSO) and distribution system operator (DSO) . By communicating this flexibility set, coordination problems, such as optimal power flow (OPF) , can be solved through a hierarchical coordination framework [ 7 ] , as shown in Fig. 1 . The primary challenge lies in the inherent nonlinearity of power system models. Existing flexibility aggregation methods often rely on linearized power flow models [ 8 , 6 ] , with the LinDistFlow model commonly used [ 9 , 10 ] . By omitting line losses, the LinDistFlow model [ 11 ] can be viewed as a linearized version of the DistFlow model [ 12 , 13 ] , which serves as an exact angle relaxation of the AC model for radial distribution. However, these linearized models present significant issues: operational points within the flexibility set derived from linearized models may violate physical constraints in the AC model, posing risks to grid stability. Figure 1 : Coordination framework between TSO-DSO In the case of the AC model, where the flexibility boundaries are nonlinear, early aggregation methods relied on the Monte Carlo approach [ 14 , 15 ] . This method generates numerous operational scenarios and tests their feasibility. While straightforward to implement, this approach is computationally intensive and unreliable due to the randomness inherent in Monte Carlo sampling, which can result in poor-quality flexibility representations. As an alternative, direction-based sampling methods have been proposed [ 16 , 17 , 18 ] . These methods optimize power at the point of common coupling (PCC) to identify extreme boundary points along specific directions, offering a clearer visualization of the flexibility region. However, they do not streamline TSO - DSO coordination, as the boundary points are not expressed analytically. This requires additional verification steps to determine whether an operational point is within the feasible region. Furthermore, the computational complexity of these methods increases with the number of sampled points, adding to the overall computation time. To balance the accuracy and computation burden, this paper focuses on improving the flexibility aggregation using the LinDistFlow model. The key contributions of this paper are as follows: \bullet Through an analysis of the LinDistFlow model, we demonstrate that the system errors omitted by the model accumulate at the coupling point, making it unsuitable for flexibility aggregation. This is illustrated through tutorial examples. \bullet We propose a compensation method to address the system errors in the LinDistFlow model. The compensated power is represented as a quadratic mapping of power exchanges within the LinDistFlow-based flexibility set. This intuitive approach is both effective and computationally efficient, while preserving privacy by preventing the disclosure of detailed distribution grid information. \bullet Beyond tutorial examples, we extend the simulations to coordinated integrated transmission-distribution (ITD) systems, focusing on the impact of electricity price variations and scenarios involving multiple distribution systems. The results validate the proposed compensation method’s ability to improve accuracy with minimal additional computational effort. The remainder of this article is structured as follows: Section II introduces the power system model and the concept of coordination using the flexibility set. Section III presents the flexibility aggregation methods based on the LinDistFlow model and the proposed system loss compensation approach, illustrated with tutorial examples. Section IV provides more extensive simulations to evaluate the proposed methods. Finally, Section V concludes the paper. II Problem Statement Consider a power grid modeled as integrated transmission-distribution (ITD) system, represented by the tuple (\mathcal{N},\mathcal{L},\mathcal{S}) , where \mathcal{N} represents the set of buses, \mathcal{L} represents the set of branches, and \mathcal{S} represents the set of subsystems, which include both transmission and distribution systems, i.e., \mathcal{S}=\mathcal{S}^{\text{T}}\cup\mathcal{S}^{\text{D}}. In this paper, we focus on a single transmission system, i.e., \left|\mathcal{S}^{\text{T}}\right|=1 , connected with multiple radial distribution systems, i.e., \left|\mathcal{S}^{\text{D}}\right|\geq 1 . We assign the transmission system an index of 1 , i.e., \mathcal{S}=\{1\} . Each subsystem \xi\in\mathcal{S} is represented as a directed graph (\mathcal{N}_{\xi},\mathcal{L}_{\xi}) , where \mathcal{N}_{\xi} and \mathcal{L}_{\xi} denote the set of buses and branches within the subsystem \xi , respectively. For every radial distribution subsystem \xi\in\mathcal{S}^{\text{D}} , the number of buses N^{\text{bus}}_{\xi} and branches N^{\text{line}}_{\xi} are related by N^{\text{bus}}_{\xi}=N^{\text{line}}_{\xi}+1 . The power generation in the distribution systems refers to the controllable power output of distributed energy resources (DERs). The transmission and each distribution system are interconnected at a point of common coupling (PCC) , typically located at distribution substations corresponding to the first bus in the local bus set \mathcal{N}_{\xi} . In this paper, we assume that the voltage magnitude at the PCC is fixed by the substations and is independent of the transmission system voltage. II-A Model of Transmission Systems In the transmission system ( \xi=1 ), the complex voltage at a bus i\in\mathcal{N}_{1} is expressed in polar coordinates, i.e., V_{i}=v_{i}e^{\theta_{i}} , where v_{i} and \theta_{i} are the magnitude and angle of the complex voltage V_{i} . Thereby, the classic AC model of the transmission system can be written as follows \displaystyle p_{i}^{g}-p_{i}^{\xi}=v_{i}\sum_{k\in\mathcal{N}}v_{k}\left(g_{ik}\cos\theta_{ik}+b_{ik}\sin\theta_{ik}\right), \displaystyle\forall i\in\mathcal{N}_{1} (1a) \displaystyle q_{i}^{g}-q_{i}^{\xi}=v_{i}\sum_{k\in\mathcal{N}}v_{k}\left(g_{ik}\sin\theta_{ik}-b_{ik}\cos\theta_{ik}\right), \displaystyle\forall i\in\mathcal{N}_{1} (1b) \displaystyle\underline{v}_{i}\leq v_{i}\leq\overline{v}_{i},\;\;\underline{p}_{i}^{g}\leq p_{i}^{g}\leq\overline{q}_{i}^{g},\;\;\underline{q}_{i}^{g}\leq q_{i}^{g}\leq\overline{q}_{i}^{g}, \displaystyle\forall i\in\mathcal{N}_{1} (1c) where p^{g}_{i} and q^{g}_{i} (resp. p^{\xi} and q^{\xi} ) denote the real and reactive power produced by generators (resp. send to distribution system \xi ) at bus i . These variables are set to 0 if no generator (resp. distribution system) is connected to a bus i . The parameters g , b denote the real and imaginary part of the complex nodal admittance matrix Y . The signs \underline{\cdot} and \overline{\cdot} denote upper and lower bounds for the corresponding state variables. II-B Model of Distribution Systems In the following, we introduce two models for the distribution system: the DistFlow model [ 12 , 13 ] and the LinDistFlow model [ 11 ]

Read the complete text