Stability analysis in a grid-interactive residential nanogrid using Markov Chains

Abstract

Statistical tools are useful in analyzing the longterm techno-economic implications in system designs. Methods such as Monte Carlo simulations and Decision Tree were applied in renewable energy system analysis due to the stochastic parameters involved. However, the methods were cumbersome and data-intensive that required lots of empirical data. Assumptions such as scenario generation in providing the required data affect quality and speed of Monte Carlo implementations. While Decision Tree tends to be cumbersome and time consuming when involved in large transitions. This paper proposed a Markov Chains method to analyze the operational stability in a photovoltaic/wind/battery residential nanogrid interacting with main grid. The proposed method only required simple states' transition probabilities that form Markovian matrices. The simulated Markovian matrices hence produced probabilistic information with several options interpreted in decision making. Results obtained indicated Markovian matrices derived from transition probabilities in nanogrid's autonomous operations and main grid interactions produced steady-state probability ratios 0.5:0.5, 0.4667:0.5333, 0.4286:0.5714, and 0.3846:0.6154. The probabilistic information indicated that the nanogrid was able to achieved 38.46-61.54% autonomy range in the lifetime analysis. The Markov Chains' performance in the nanogrid/main grid energy trade-offs is envisaged to be improved by considering each transition state supplementing one another.