In remote or islanded communities, the use of microgrids (MGs) is necessary to ensure electrification and resilience of supply. However, even in small-scale systems, it is computationally and mathematically challenging to design low-cost, optimal, sustainable solutions taking into consideration all the uncertainties of load demands and power generations from renewable energy sources (RESs). This paper uses the open-source Python-based Energy Planning (PyEPLAN) tool, developed for the design of sustainable MGs in remote areas, on the Alderney island, the 3$^rd$ largest of the Channel Islands with a population of about 2000 people. A two-stage stochastic model is used to optimally invest in battery storage, solar power, and wind power units. Moreover, the AC power flow equations are modelled by a linearised version of the DistFlow model in PyEPLAN, where the investment variables are it here-and-now decisions and not a function of uncertain parameters while the operation variables are it wait-and-see decisions and a function of uncertain parameters. The $k$-means clustering technique is used to generate a set of best (risk-seeker), nominal (risk-neutral), and worst (risk-averse) scenarios capturing the uncertainty spectrum using the yearly historical patterns of load demands and solar/wind power generations. The proposed investment planning tool is a mixed-integer linear programming (MILP) model and is coded with Pyomo in PyEPLAN.