This paper addresses a novel probabilistic optimisation framework for handling power system uncertainties in the optimal power flow (OPF) problem that considers all the essential factors of great impact in the OPF problem. The object is to study and model the correlation and fluctuation of load demands, photovoltaic (PV) and wind power plants (WPPs) which have an important influence on transmission lines and bus voltages. Moreover, as an important tool of saving waste heat energy in the thermoelectric power plant, the power networks share of combined heat and power (CHP) has increased dramatically in the past decade. So, the probabilistic OPF (POPF) problem considering valve point effects, multi-fuel options and prohibited zones of thermal units (TUs) is firstly formulated. The PV, WPP and CHP units are also modeled. Then, a new method utilizing enhanced binary black hole (EBBH) algorithm and 2m+1 point estimated method is proposed to solve this problem and to handle the random nature of solar irradiance, wind speed and load of consumers. The correlation between input random variables is considered using a correlation matrix. Finally, numerical results are presented and considered regarding the IEEE 118-busses, including PV, WPP, CHP and TU at several busses. The simulation and comparison results obtained demonstrate the broad advantages and feasibility of the suggested framework in the presence of dependent non-Gaussian distribution of random variables.