Since the number of loops in refineries or petrochemical plants is very large and the number of loops with poor performance is equally large, to prioritize their maintenance is essential to ensure plant profitability. This work proposes a methodology called LoopRank to compute the importance factor of each loop, aiming to prioritize their maintenance. The algorithm is based on the connection among them, which is computed using partial correlation. The algorithm is based on PageRank, which analyses connections among nodes recursively and computes a rank for each node using partial correlation. The LoopRank assigns an individual score for each loop ranging from 0% to 100%. Based on this score, the loop maintenance can be ranked. The LoopRank algorithm is computationally efficient, thus allowing its industrial large-scale application. The proposed algorithm was applied both on simulation and industrial case studies, providing fruitful results.