Large Scale CO2 Releases for Dispersion Model and Safety Study Validation

Abstract

Abstract The second phase of the CO2PIPETRANS Joint Industry Project (JIP) aims to fill knowledge gaps associated with the safe and reliable pipeline transport of CO2. The JIP has three main focus areas and technical work packages designed to address these. The work packages collect experimental data and experience on dense phase CO2 release model validation data, pipeline fracture arrest, and corrosion. This paper presents an overview of results and conclusions from the work package focusing on collection of data for validation of release models. The JIP consists of 15 partner organisations, who are: Arcelor Mittal, BP, DNV GL, Endesa, ENI, E.on Ruhrgas, Gassco, Gassnova, Health and Safety Executive (HSE) UK, Maersk Oil, Petrobras, Petroleum Safety Authority (PSA) Norway, Shell, V&M Tubes, and Vattenfall. The objective of the CO2 release model validation data work package was to collect and make available data for validating release and dispersion models for dense phase CO2. This involved releasing dense phase CO2 from large inventories through orifices ranging in size 6 mm to 150 mm onto a suitably sized array of instruments to measure the CO2 concentration and temperature profiles. The condition of the CO2 being released was up to 150 barg and 150 °C with release durations up to 10 minutes. Most of the releases were in a horizontal orientation 1m above the ground but during some of the tests the orientation was changed to be upwards, downwards to impact the ground, or into an enclosure. In addition to the CO2 releases described above, the work package also undertook rapid depressurisation of a long horizontal pipe containing 100 barg CO2 in order to collect data, amongst other aspects, on shock wave propagation. These experiments used a 200m long, 50 mm diameter pipe mounted on load cells with an orifice plate and explosive initiated bursting disk at one end. In addition to measuring data within the pipe during the rapid depressurisation tests, dispersion and temperature data was also recorded downstream of the release for release model validation purposes. The paper presents an overview of the frontier experimental work mentioned above along with discussion on the output of the data review that was subsequently completed during which recorded data was compared with model predictions. Validation of dispersion models and safety studies reduces the uncertainty and hence conservatism required in these studies thereby making design and implementation of CO2 pipelines safer and more cost effective.