In the present study, the models of flow, heat transfer and reforming of 2-D single reforming tube and 3-D reformer core inside the reformer were implemented by CFD. The models were evaluated by comparison of simulations with data derived from a catalytic packed-bed reactor. In addition, a NG engine-reformer experimental system was established. Thereby, the experimental studies on performance of the reformer and engine under different excess air coefficient (λ), exhaust gas recirculation rate (XEGR) and reformed NG supply (CH4/O2, M/O value) were conducted based on IMO Tier III and China II emission standards. The results showed that oxidation and steam reforming reactions of methane were dominant in the process of exhaust gas-methane reforming and had obvious reaction sequence. For the reformer optimization, the design parameters were determined: GHSV = 16,000 h-1 and L/D = 7 for single reforming tube; diffuse angle of 30° and one deflector for reformer core. Besides, for experiments of REGR system, the reformer should operate under the condition of near M/O = 2-2.5 and XEGR = 7.3 % with high λ to achieve an optimal H2 yield and efficient H2 production. Compared with original engine under 50 % load and XEGR = 10.4 %, the brake thermal efficiency of the engine with REGR was increased by the maximum of 0.7 % (λ = 1.31). Noted that the maximum 85.5 % reduction of HC + NOx and the minimum 64.1 % treatment rate of CH4 could be obtained for meeting China II at λ = 1.25. Under different propeller loads, HC + NOx and CO emissions could both satisfy IMO Tier III and China II standards via adjusting the matched strategy of λ and XEGR. To meet fully China II standard, the minimum treatment rate of CH4 by the post-treatment unit needed to reach 79 %, 64 % and 64 % at 25 %, 50 % and 75 % load, respectively.
Keywords: Emission standards; Marine NG engine; Reformer; Simulation and experiment; Structural optimization.
Copyright © 2023 Elsevier B.V. All rights reserved.