In this article, we discuss the uncertainties involved in the models and in the measurements necessary to estimate the emissions of greenhouse gases (GHG) in a chemical industry. When these uncertainties cannot be neglected and some measurements exhibit correlations with others, estimating the final emission is not a trivial task. Even if we intend to determine a simple point estimate for the mean emissions, we will need to use the average values of the measurements as well as information about their uncertainties and correlations in complex computations. To solve this problem, we propose a Monte Carlo method to estimate the mean and confidence interval of CO₂ emissions in the context of uncertainties and correlations. We validated our approach through a case study involving a traditional chemical company in Brazil. Our results indicated that previously, there was an overestimation of the emission because the company did not consider uncertainties and correlations. The overestimation was modest since the parameters involved present relatively little uncertainty, but the bias effect was clear. This research has demonstrated the importance of accounting for uncertainties and correlations in emission estimates, providing a practical framework for analyses in industrial settings.