Proposal of a modified algebraic approach to evaluate required capacity for load frequency control (LFC) under a large penetration of solar photovoltaic generations

Currently, power balancing in a power system is achieved through economic dispatch control (EDC) and load frequency control (LFC). Under a high penetration of solar photovoltaic (PV) generations, forecasts of long-period fluctuations in PV generation output are incorporated into EDC, and additional LFC capacity is necessary to compensate for forecast error and short-period fluctuations in PV generation output. Forecast of PV generation output is much more difficult compared to load forecast, and forecast error may be large. It is important to take such characteristics of PV generation output forecast error into account when evaluating required LFC capacity under a high penetration of PV generations. Until now, an algebraic approach has been used to evaluate required LFC capacity. This approach does not take into account forecast error for long-period fluctuations in PV generation output. In this paper, a modified algebraic approach is proposed, which is suitable for assessing required LFC capacity under a high penetration of PV generations. The following modifications to the conventional method make it possible to evaluate required LFC capacity against not only short-period fluctuations in demand and PV generation output, but also against long-term deviation caused by forecast error for PV generation output. (1) Period of fluctuation to be analyzed is expanded to long-term and algebraic approach is applied to specific ranges of period concerned. (2) Utilization of equivalent residual vector composed of two orthogonal vectors - residual of power balancing control and forecast error of PV generation output - makes it possible to factor forecast error into the assessment. To show the effectiveness of the proposed method, an estimation of required LFC capacity is made for a 10GW model system on a holiday in May. The results show that long-term PV generation output forecast error may have a great impact on required LFC capacity, which cannot be assessed using a conventional algebraic approach.