Environmental awareness and energy policies led to decarbonization targets, fostering the adoption of distributed energy resource in the distribution network. Particularly, photovoltaic systems have been gaining momentum due to cost-competitive option and financial benefits. However, traditional distribution networks were not designed for intermittency in power generation. This poses technical issues such as reverse power flow, overvoltage, and thermal overloading. Furthermore, the growth in intermittency and variability of distributed energy resources increases the uncertainty, hence, it brings challenges for the operation, planning, and investment decisions. In this context, probabilistic methods to cater for these uncertainties are essential to address this issue. This paper presents a probabilistic power flow method based on point estimate method combined Edgeworth expansion for high penetration of photovoltaic generation in distribution networks. Normal distribution and Beta distribution are considered for load and solar irradiation modelling, respectively. The method is assessed for different cases using the IEEE 33-bus distribution test system with photovoltaic systems installation. The point estimate method combined Edgeworth expansion provided satisfactory results with lower computational effort and high fitting accuracy of statistical information compared to Monte Carlo simulation.