Towards proactive control of grid-scale battery energy storage using dynamic preview length

The finite preview of energy price forecasts or the pre-dispatch price has potential to improve performance of batteries participating in energy market and therefore must be included as a key control design parameter. Fixed preview and reactive (one-step ahead) preview strategies are often suboptimal and computationally inefficient because, in such approaches, the batteries are exposed to the uncertain pre-dispatch issued by the market operator. In addition, energy market price volatility changes across seasons and times of day, so a finite preview that is optimal in one period may perform poorly in another. This paper proposes a dynamically adjusted preview length for proactive operation of grid-scale batteries, particularly Community Battery Storage (CBS). A supervisory layer uses Bayesian Online Change-Point Detection (BOCPD) to track real-time pre-dispatch prices, detect regime shifts, and tune the preview horizon. A lower-layer RHC then updates dispatch decisions every five minutes using the adjusted horizon, while enforcing time-varying dynamic operating envelopes for network-aware operation. By adapting its preview length to market volatility, the proposed method significantly increases battery revenue across representative daily scenarios and achieves an overall 9.23% year-round revenue improvement compared with the conventional fixed-preview approach, while also maintaining lower computational burden.