Effect of SO 2 , O 2 , NO 2 , and H 2 O concentrations on chemical reactions and corrosion of carbon steel in dense phase CO 2 . Bjørn H. Morland †,1,2 , Truls Norby 1 , Morten Tjelta 2 , and Gaute Svenningsen 2 Article history: Received Day Month Year Accepted Day Month Year Available Day Month Year Keywords: A. CO2 corrosion B. Iron sulphate C. Dense phase CO2 D. Impurities E. NO2 F. SO2 1 University of Oslo, Department of Chemistry, FERMiO, Gaustadalléen 21, NO- 0349 Oslo, Norway. 2 Corrosion Technology department, Instituttveien 18, 2007 Kjeller, Norway. †Corresponding author: Email: bjorn.morland@ife.no. ABSTRACT Carbon capture, utilization and storage (CCUS) is expected to be important method for reducing the CO 2 emissions to prevent global warming. Several species could follow the CO 2 through the capture plant as carry over. It is expected that nitrogen dioxide (NO 2 ), sulphur dioxide (SO 2 ), oxygen (O 2 ), and water (H 2 O) can be present as impurities (concentrations at the ppmv level) in the captured CO 2 . The exact composition will depend on the flue gas type, the CO 2 capturing process and multiple other parameters. Some of these impurities are reactive and may cause corrosion in carbon steel pipelines and could therefore be a threat for safe CO 2 transport. The present study used a novel experimental setup to realistically simulate a CO 2 transport pipeline system with a controlled and variable concentration of impurities at a total pressure of 100 bar and a temperature of 25 °C. The water concentration was increased and decreased with constant concentration of SO 2 and O 2 , to observe and identify possible reactions or threshold levels which could give corrosion. A similar experiment was conducted with NO 2 . First, experiments were carried out without steel coupons, to observe un- catalysed reactions, and then with coupons to measure corrosion rates. The first sign of corrosion appeared at 350 ppmv of water with NO 2 present. At 670 ppmv water with 75 ppmv NO 2 the overall corrosion rate was about 0.57 mm/y and the main product was iron oxide. The corrosion process for SO 2 , O 2 , and water was much slower, and the first sign of corrosion appeared around 1900 ppmv of water, with about 75 ppmv of SO 2 and 230 ppmv of O 2 . The corrosion rate increased some when the water concentration was increased to 2400 ppmv, but the overall corrosion rate was only 3.6 µm/y and the main product on the surface was iron sulphate. INTRODUCTION Carbon capture, utilization and storage (CCUS) is needed to meet the goal set by the International Energy Agency (IEA) 1 and the Intergovernmental Panel on Climate Change (IPCC) 2 of limiting the long- term global temperature rise. It will be essential that the cost of the CCUS processes is kept as low as practically possible to achieve the goal. For this reason, carbon steel is most economically feasible material for long transport pipelines. Even if carbon steel is an attractive material for constructing pipelines for transport of dense phase CO 2 , not to mention network of pipelines already in place from oil and gas production, it may corrode if the captured CO 2 contains certain combinations of additional species (impurities), especially liquid water. If the water is completely dissolved in the dense phase CO 2 , the corrosion rate would be very low, typically around 1 µm/y. 3 Presence of certain species together with water, may increase the corrosion rate and change the type of corrosion products. In the suggested specifications for CO 2 transport, 4, 5 the maximum allowed limit for several impurities like NO 2 , SO 2 , and H 2 S were set for health, safety, and environmental reasons, without considering possible corrosion reactions. Previous work by the authors has shown that water, NO 2 , SO 2 , H 2 S, and O 2 impurities can react and create an aqueous phase that contains sulphuric and nitric acid. 6 Even if most CO 2 transport system will not have liquid water present, some water is expected to be present at a concentration well below the solubility limit. This low level of water is probably sufficient to create a thin surface layer (some monolayers) of water at the metal surface. It is known that both SO 2 and NO 2 dissolve in a liquid water phase to form acid. However, reactions in bulk water phase may be different from reactions in thin water films, and it is not known if these acids can form in thin surface layers or if there is a critical layer thickness before the reactions can occur (like in atmospheric corrosion where the adsorbed water reaches the properties of bulk water when the relative humidity exceeds 60 – 70% 7 ). Corrosion and chemical reactions in dense phase CO 2 have been investigated in some experiments with single or multiple impurities. Halseid et al. summarized most of the experiments up to 2014 in a review. 8 There is only a limited number of papers addressing CO 2 -H 2 O- NO 2 systems, but Dugstad et al. 9 reported corrosion rates of carbon steel from 0.2 up to 1.7 mm/y, depending on the concentration of dissolved water. It was however emphasized that the corrosion rate could be higher due to consumption of the impurities in the corrosion process, meaning that the concentration of the impurities was significantly lower or zero in the end of the experiments. Paschke et al. executed a series of experiments with multiple impurities among other NO and they conclude that the corrosion rate was lower than 80 µm/y as long as the water concentration was less than 1000 ppmv. 10 Several papers address CO 2 -H 2 O-SO 2 systems, and wide range of corrosion rates from 0.005 to 7 mm/y have been reported 11-20 depending on the concentration of impurities, temperature, flow, and exposure time. Hua et al. 19 indicated that the critical water content with SO 2 /O 2 present should be no higher than about 500 ppmv (mole) to minimize localized corrosion attacks, while for avoiding general corrosion rates in excess of 0.1 mm/y the water content should be less than 1900 ppmv. These findings are in