Hydrogen energy can be effectively converted from various energy sources, such as renewable energy or hydrocarbon fuel, and therefore, it is a promising energy source. Various hydrogen production processes have been proposed worldwide because conventional energy conversion systems, after minor design modifications, can be used for such conversions. However, hydrogen gas can be more dangerous than other flammable gases if released into the atmosphere, where it can rapidly reach its explosion limit because of its high diffusion velocity. Therefore, a failure mode and effects analysis (FMEA) for hydrogen production processes is required. In this study, FMEA is performed for hydrogen produced from coal syngas using Thermococcus onnurineus NA1, which was discovered in the Indonesian deep sea. A fuzzy inference methodology is introduced for a quantitative analysis of the linguistic ambiguity of risk priority number estimated from a conventional FMEA. To estimate the objective severity ranking, we introduced the potential asset loss combined with the proportionate cost of each piece of equipment under total capital investment and a weight value influenced by environmental and mankind. Moreover, we proposed a fuzzified risk matrix to effectively represent the fuzzy risk priority number (f-RPN) under the risk matrix; variation with and without fuzzification of risk priority number is then expressed to ascertain why this variation has occurred through a vector diagram. Based on the f-RPN vector diagram, we have performed a design revision of the hydrogen production process using Thermococcus onnurineus NA1 to adjust the risk priority number downward from a broadly unacceptable region. (C) 2019 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.