Variable reactive power compensators versus fixed and switchable capacitor banks in providing ancillary services considering voltage fluctuations

Voltage fluctuations across distribution network buses can significantly influence both load consumption patterns and the reactive power output of compensation devices. To address these effects, strategic placement of reactive power support equipment is essential. While prior research has extensively examined the positioning of capacitors for loss minimization and voltage enhancement, limited attention has been given to the economic implications of deploying variable reactive power compensators—such as thyristor-switched capacitors (TSCs) or static VAR and static synchronous compensators (SVCs and STATCOMs)—for ancillary service provision under dynamic voltage conditions. Moreover, representative analytical approaches may rely on nonlinear optimization techniques or simplified linear models, creating a tradeoff between model fidelity and computational effort when the sensitivity of reactive power output to voltage variations. Nonlinear approaches can be computationally intensive and difficult to implement. In this paper, a computationally efficient mixed-integer second-order cone programming (MISOCP) framework is proposed to incorporate voltage fluctuations into the economic evaluation of fixed and switchable capacitors, and variable reactive power devices. The model is solved using commercial linear solvers and enables effective planning of reactive power resources while maintaining computational tractability. Case studies on standard 33- and 69-bus distribution systems declare that the proposed approach improves technical and economic indicators in the investigated cases. In the 33-bus system, variable reactive power compensators reduced total annual cost by 76.3% compared with fixed capacitors while improving the minimum voltage from 0.93 to 0.981 pu and generating annual ancillary service revenues of up to 13,578$. In the 69-bus system, variable reactive sources achieved up to 13.2% greater cost savings and 11.7% higher loss reduction than fixed capacitors. The results further show that voltage-dependent load modeling significantly affects the optimal allocation and sizing of reactive compensation devices, with constant-impedance loading conditions generally yielding the lowest losses and best voltage profiles. Furthermore, the proposed framework obtained high-quality solutions within only a few seconds, demonstrating its computational efficiency. These findings demonstrate, for the investigated systems, the potential technical and economic benefits of integrating voltage-dependent load and reactive power behavior, flexible reactive power resources, and ancillary-service revenues into distribution system planning.