Exposing Barriers to Flexibility Aggregation in Unbalanced Distribution Networks

The increasing integration of distributed energy resources (DER) offers new opportunities for distribution system operators (DSO) to improve network operation through flexibility services. To utilise flexible resources, various DER flexibility aggregation methods have been proposed, such as the concept of aggregated P-Q flexibility areas. Yet, many existing studies assume perfect coordination among DER and rely on single-phase power flow analysis, thus overlooking barriers to flexibility aggregation in real unbalanced systems. To quantify the impact of these barriers, this paper proposes a three-phase optimal power flow (OPF) framework for P-Q flexibility assessment, implemented as an open-source Julia tool 3FlexAnalyser.jl. The framework explicitly accounts for voltage unbalance and imperfect coordination among DER in low voltage (LV) distribution networks. Simulations on an illustrative 5-bus system and a real 221-bus LV network in the UK reveal that over 30% of the theoretical aggregated flexibility potential can be lost due to phase unbalance and lack of coordination across phases. These findings highlight the need for improved flexibility aggregation tools applicable to real unbalanced distribution networks.

💡 Research Summary

The paper tackles a critical gap in the emerging field of distributed flexibility services: the discrepancy between theoretical aggregated P‑Q flexibility potential and what can actually be delivered in real low‑voltage (LV) distribution networks that are inherently unbalanced. While many recent studies propose aggregation methods based on P‑Q flexibility areas, they typically rely on single‑phase power‑flow models and assume perfect coordination among all distributed energy resources (DER). The authors identify two concrete barriers that can dramatically shrink usable flexibility.

Barrier 1 – Lack of DER coordination. Most aggregation frameworks treat flexible units as centrally controllable, allowing simultaneous adjustments across all phases to satisfy voltage constraints. In practice, DER are often single‑phase, behind‑the‑meter, or suffer from limited observability and communication, making coordinated control across phases difficult or impossible.

Barrier 2 – Phase‑unbalance effects. LV networks frequently exhibit significant voltage unbalance, quantified by the voltage unbalance factor (VUF). Single‑phase models ignore VUF limits, whereas real networks must keep VUF below regulatory thresholds, constraining how much active and reactive power can be injected on each phase.

To quantify the impact of these barriers, the authors develop a three‑phase nonlinear AC optimal power flow (OPF) formulation that explicitly includes (i) complex voltages and power injections per phase, (ii) the positive‑ and negative‑sequence voltages and VUF constraints for a selected set of buses, and (iii) phase‑coordination constraints that fix the power outputs of DER on certain phases, thereby simulating imperfect coordination. The objective function is a weighted combination αp · Pref + αq · Qref, where the coefficients αp and αq are varied to trace the boundary of the aggregated P‑Q flexibility region at a chosen reference bus. By iteratively solving the OPF for different (αp, αq) pairs, the method constructs the feasible P‑Q envelope.

The mathematical model is implemented in Julia, extending the open‑source PowerModelsDistribution.jl library, and released as the open‑source tool 3FlexAnalyser.jl. Two case studies are presented: an illustrative 5‑bus feeder and a real 221‑bus LV network from the United Kingdom. Both are examined under four scenarios: (1) ideal – perfect coordination, no VUF limits; (2) only VUF limits; (3) only coordination limits; and (4) both VUF and coordination limits.

Key findings from the simulations are:

1. Voltage unbalance alone reduces the theoretical flexibility by roughly 15 % because the VUF constraint forces the optimizer to keep phase voltages within tight bounds, limiting simultaneous active‑reactive injections.
2. Lack of coordination alone leads to a loss of about 25 % of flexibility, as the optimizer cannot exploit cross‑phase compensation to alleviate voltage violations.
3. Combined barriers are most detrimental: over 30 % of the theoretical aggregated flexibility becomes infeasible. In many Monte‑Carlo runs, uncoordinated DER cause VUF violations in more than 15 % of scenarios.
4. Small single‑phase units can collectively provide more flexibility than a few large three‑phase units, but only if they are perfectly coordinated. Without coordination, they tend to exacerbate existing unbalance.
5. Worst‑case conditions occur when tight VUF limits coexist with severe coordination restrictions, rendering the OPF infeasible for a large portion of the P‑Q space.

The paper contributes (i) a novel framework that merges three‑phase OPF with the concept of aggregated P‑Q flexibility areas, (ii) the first quantitative assessment of how phase‑unbalance and coordination constraints shrink these areas, and (iii) an open‑source toolbox that enables other researchers and system operators to replicate and extend the analysis.

Beyond the technical results, the authors discuss practical implications: system operators must consider the cost of installing advanced metering, communication, and control infrastructure to achieve the necessary coordination; market designs should internalize the “coordination penalty” and “unbalance penalty” so that flexibility bids reflect the true deliverable capacity; and planners should prioritize balanced phase allocation of new DER to mitigate VUF growth.

Future research directions suggested include stochastic flexibility region computation that incorporates forecast uncertainty, distributed optimization techniques for large‑scale networks, and the integration of dynamic (time‑varying) VUF limits reflecting evolving load patterns. Overall, the study convincingly demonstrates that ignoring unbalance and coordination realities can lead to overly optimistic assessments of DER flexibility, and it provides a rigorous, reproducible methodology to obtain realistic flexibility envelopes for unbalanced LV distribution systems.