The sizing of equipment spares is an important investment analysis for the design and planning of the operation and maintenance of any large system. These concepts can also be applied to electric distribution systems of the interconnected Brazilian network. Distribution systems, regulated by the Brazilian Electricity Regulatory Agency, need to supply energy in a secure, stable, and extremely low fault rate, making monetary penalties possible in case of non-compliance with established standards. One of the most sensitive points in a distribution system is its current substations, where failures impact a large number of consumers, besides having relatively high-cost equipment. Redundancy in systems of this type could considerably reduce failures, but would require a very high initial investment. Thus, the need of a strategic inventory becomes relevant, aiming to reduce the duration of the failures with the least possible investment. Transformers are of great importance in an electric power station, whose spare requirements are usually assessed, but would only serve a locality or region, while the stock of mobile units, although on a temporary basis, would serve to different affected regions. This dissertation presents a probabilistic methodology for assessing of spare transformer requirements and the associated inventory acquisition strategy over time, focusing on mobile units. It includes the evaluation of reliability indices and comparison of investment and operational costs for different scenarios. It also allows important issues to be considered, such as equipment aging, load increasing with different characteristics for each unit, and the possibility of load transfers from one faulty unit to another in operation. These assumptions provide more accurate input data for investment analysis, considering the individual conditions of equipment. The basic tool to be used is the chronological Monte Carlo simulation that accurately captures all characteristics of the problem being solved. The proposed methodology is applied to a group of transformers of the class 138-13.8 kV, 25 MVA, aiming to demonstrate the ability of the proposed method to find the appropriate solutions from a technical and economic points of view.