State estimators, used to monitoring transmission networks, are not widely applied to distribution systems. Before deploying such a system, DSOs want to evaluate its performances in field conditions. In this paper, different uncertainty sources are considered. Device measurement error is modelled by a normal distribution function, and measurement delays or loss of communication are emulated by simulations. This paper introduces a tool to evaluate the performance of a state estimator, applied on a 102-node real LV network of the Swiss DSO Romande Energie, with 40 smarts meters and three GridEye monitoring devices, a new GPS synchronized high precision measurement device. The method consists in the comparison between load flow calculations and an estimator using a weighted least squares formulation. Uncertainties are added to the load flow results and a selection of them are considered as measurements. Thereby, the performance is evaluated for different simulation cases: meters placements, GridEye cells placements and different datasets for each measurement device can be selected. The simulations are based on several snapshots representing different times of the day.