Petrochemical companies are prone to fire accidents. In order to study the fire risk coupling of petrochemical companies, the traditional N-K model combined with the analytic hierarchy process is used to replace the accident probability with the frequency of risk factors to calculate the risk coupling degree. Research shows that the improved N-K model makes up for the limitation of the traditional N-K model that requires a large amount of data support; the more coupled risk factors, the greater the risk value; human factors must be paid attention to when managing petrochemical companies.