Comprehending and assessing the vast amount of information within technical specifications, procedure manuals, and work orders associated with nuclear power plants is challenging. The demands associated with this task are particularly evident during power plant outages, which create a pressing need for quickly and effectively identifying the right information at the right time for making the right decision. Text mining methods that use natural language processing (NLP) approaches to automatically extract and store document information might assist managers of power plant outages by reducing the cognitive load. NLP can parse unstructured text in documents and convert it into structured form for storage of conceptual information and relationships. We are exploring the feasibility of using NLP for helping in outage management. As a proof-of-principle, we are designing an NLP tool to identify constraints in the configuration of components during the procedure used to test a low-pressure safety injection (LPSI) pump. The NLP tool is extracting action verbs that represent instructions for changing the state of power plant components. The extracted information will form the analytical backend for determining whether proposed changes in configuration lead to constraint violations, which will be displayed via a graphical user interface.