We discussed the models and methods used to detect anomalies in the functioning of electric power systems, a vivid example of which are the cascade accidents. The normal mode of operation of these systems can be described as the spectral decomposition of the virtual energy of the system, which can be represented as the spectral decomposition of the squared H2 -norm of the transfer function for fixed input and output of the system. When approaching the emergency mode of operation, the virtual energy balance changes. The share of energy of weakly stable modes increases, which makes it possible to identify the cause of the imbalance: approaching one or several the network cutsets to the limits of its capacity. One can make a cautious assumption that the method of gramians can be used to describe changes in the balance of energy in a human body. One can formulate a cautious hypothesis. For example, the method of gramians can be used to assess the risk of developing diabetes of the first or second types by means of anomalies detecting in the spectral decomposition of the virtual energy of the pancreas and / or its organs. Confirm or disprove the hypothesis will help further researches.