Multi-period planning of distribution grid reinforcements under uncertainty about future penetration of photovoltaic systems

The policy transition towards renewable sources of energy generation has triggered a proliferation of small, independent, and decentralized power plants such as solar PV panels owned by households. This presents network operators with challenges regarding the maintenance and expansion of low-voltage distribution grids. We model multi-period planning of grid reinforcement on the basis of three investment strategies: a purely reactive strategy that disregards any information about the future, a proactive strategy that requires only a probability distribution of future distributed generation, and a hypothetical strategy assuming perfect information. Simulations show that 75% of the time the proactive strategy achieves equal or lower total reinforcement cost than the reactive strategy, which can be regarded as the industry standard. On average, it yields about 30% of theoretically possible savings. The results are robust to changing simulation parameters, including PV penetration levels.