We present a statistically enhanced version of the GreenHouse gas emissions of current Oil Sands Technologies model that facilitates characterization of variability of greenhouse gas (GHG) emissions associated with mining and upgrading of bitumen from Canadian oil sands. Over 30 years of publicly available project-specific operating data are employed as inputs, enabling Monte Carlo simulation of individual projects and the entire industry, for individual years and project life cycles. We estimate that median lifetime GHG intensities range from 89 to 137 kg CO2eq/bbl synthetic crude oil (SCO) for projects that employ upgrading. The only project producing dilbit that goes directly to a refinery has a median lifetime GHG intensity of 51 kg CO2eq/bbl dilbit. As SCO and dilbit are distinct products with different downstream processing energy requirements, a life cycle assessment ("well to wheel") is needed to properly compare them. Projects do not reach steady-state in terms of median GHG intensity. Projects with broader distributions of annual GHG intensities and higher median values are linked to specific events (e.g., project expansions). An implication for policymakers is that no specific technology or operating factor can be directly linked to GHG intensity and no particular project or year of operation can be seen as representative of the industry or production technology.