In the heavily industrialized western India, on the express way between Mumbai and Pune there is a series of five 03-lane one way tunnels. Risks in the tunnels on the hilly express way are sourced from the heavy goods vehicles often carrying dangerous goods. During the design stage, risk assessment is carried out using theoretical risk models such as OECD/PIARC DG-QRAM. However, during operational stage of a tunnel real time accident data are available to estimate the risk or fatality rate. In this paper a methodology has been developed to estimate aggregated risk of the series of 05 tunnels from the accident rate by Monte-Carlo simulation using simple random sampling technique. It has been found that major accident rate (or risk) is better represented by lognormal distribution.