Abstract-With the development of a new smart load device, the electric spring (ES), the operation of distribution systems with high renewable penetration becomes more flexible. The ESs can be installed at non-critical loads for grid support. This paper proposes a two-level voltage management scheme to optimize the voltage profiles of the network, especially at chosen critical buses. In the upper level, the tap positions of load tap changer and capacitor banks switching are optimized to prevent the voltages along the feeder from being out of limits. The model predictive control technique is applied to handle the uncertainties in renewable energy and demand. In the lower level, the responsibility of maintaining the voltages of the critical buses is shared among the electric springs (
II. INTRODUCTIONN the past, distribution systems have been designed for a one-way flow of energy from substations to customers. However, in recent years, increasing numbers of distributed generators have being connected at the customer side, such as wind or solar, due to environmental and economic concerns [1]. These renewables, besides offering sustainability solutions to energy crisis, can defer network reinforcement, reduce energy transaction, and save power losses. However, due to the high penetration level, energy now can flow in reverse directions during peak generation time, which is not allowed in the traditional systems, causing the voltage rise issue. Additionally, the voltage fluctuation along feeder caused by the intermittent nature of renewable energy is unacceptable for some critical energy consumers, who require more reliable energy supply with higher power quality.In a conventional distribution system without renewable energy penetration, voltage regulation is achieved through compensating voltage drop along feeder by typical means, such as load tap changer (LTC) and capacitor banks (CB). Such voltage regulation devices can respond to the voltage deviations according to different control strategies. Classic control algorithms and device models were proposed to keep the system voltages within constraints in [2]. However, in future distribution systems with the emerging voltage issues, LTC and CB are not flexible enough, especially for the voltage fluctuation mitigation [3]. The high intermittency of renewable energy causes increased number of tap operation actions and capacitor switching, which will decrease the lifetime of such devices [4], [5]. Meanwhile, the most common problem is shifted from voltage drop to voltage rise which cannot be handled by CBs. As a result, developing new control devices and control approaches for voltage regulation in a centralized or a distributed way are needed