[en] Oxygen, present at high concentrations in water, is the main cause of corrosion in process equipment such as steam generators in power production and water-cooled stator windings in turbine generators. Thus, mitigating corrosion involves removal of oxygen by a method such as mechanical deaeration, chemical scavenging or catalytic recombination with hydrogen. Where hydrogen is already present in the water or that it can be conveniently provided, the catalytic recombination has become a preferred method because of its ability to remove oxygen to very low levels (a few ppb levels) without producing any undesirable by-products. Palladium supported on anion or cation exchange resins is the most common catalyst used in industry. Canadian utilities use sulfite-based resins to scavenge the oxygen in stator cooling water application. In the 1980's, AECL demonstrated the application of wetproofed catalysts for dissolved oxygen removal under an EPRI/AECL contract. The work reported here is focussed on further development of palladium- and platinum-based catalysts on inert styrenedivinylbenzene (SDB) polymer support for dissolved oxygen removal applications. The inert nature of the catalyst support is expected to be an additional benefit for this application. A lab-scale plug flow reactor, 19-mm diameter by 100-mm long, was used for the tests at water fluxes ranging from 400 to 900 mol·s^-1·m^-2 (4.3 to 9.6 bed volumes per minute) at ambient temperatures. The catalyst performance was characterized in terms of conversion efficiency based on dissolved oxygen concentrations at the inlet and outlet of the reactor and benchmarked against the commercial catalyst Lewatit OC1045 (marketed by Bayer). The effect of the type of precious metal (platinum or palladium), the catalyst support particle size distribution, the catalyst metal loading and the molar ratio of reactants (oxygen and hydrogen) on conversion efficiency was studied to optimize the catalyst performance. Also, the effect of the presence of boron, which is used as a corrosion inhibitor and poison for the nuclear reaction, on conversion efficiency was investigated. Dissolved oxygen concentrations as low as 13 ppb were measured at the lowest water flux used for the tests, and the catalyst performance profiles in terms of conversion as a function of water flux showed the ability to reduce dissolved oxygen concentration even further by increasing the residence time (decreasing water flux for a given bed size or increasing the bed size for a given water flux). Dissolved oxygen removal efficiency in excess of 97% over the complete range of water flux was achieved. An AECL proprietary catalyst with particle size distribution similar to that of the commercial catalyst demonstrated superior oxygen removal performance, but produced higher pressure-drop. The catalytic performance was strongly dependent on the particle size distribution of the catalyst, pointing to the need to give careful consideration to pressure-drop at the design stage. The recombination efficiency increased significantly when the dissolved hydrogen concentration was increased from stoichiometric to twice stoichiometric requirement for the reaction. Since the hydrogen required for the recombination is conveniently present in the stator cooling water through ingress from the generator hydrogen system, the catalytic reactor can be located in a suitable position in the cooling water recycle loop. Alternately, the sulfite resins used for deoxygenation may be replaced with the noble-metal catalyst, which will necessitate hydrogen addition to the water. For make-up water deoxygenation, the catalyst may be incorporated as part of the system with hydrogen addition. Boron at concentrations up to 1000 ppm had no major poisoning effect on AECL catalyst performance. A design equation for deoxygenation is derived from a simple analysis of catalyst performance. (author)