A two-stage management strategy for the control of distributed storage in a low voltage network

This paper presents a management strategy for resources “behind-the-meters” along with preliminary field tests on an actual distribution grid. The overall objective is to manage a pool of distributed assets (e.g. storage systems), located in the user’s premises to satisfy metrics at the networks scales (e.g. reduced transformer loading and backfeed power). The proposed operational strategy relies on a two-stages decision making process. A first look-ahead step optimizes the storage schedule to smooth the distribution transformer power profile and is based on daily forecast for the load and renewable (solar) profiles. Then, a near real-time stage at is run locally at the scale of every user, based on actual measurements. The objective of that step is to compute the storage charge/discharge setpoints in order to fulfil the day-ahead commitment. Results on a theoretical case study are displayed. Then, a practical implementation of the control architecture is presented before the illustration of preliminary field test (for the first step of the control). Those tests were carried out on a distribution grid in Germany with a great penetration of solar panels “behind-the-meter” and a dozen households equipped with storage systems, two of which were controlled.