Predictive Monitoring and Prescriptive Maintenance to Help Decision Making in Refinery Turnarounds

Predictive monitoring and prescriptive maintenance to help decision making in refinery turnarounds

Emilio Diaz-Bejarano¹, Manuel Martinez-Tafalla², Guillermo Galan¹ and Francesco Coletti^1,*

¹Hexxcell Ltd., Foundry Building, 77 Fulham Palace Road, London, W6 8AF, UK

²Repsol C.I. Cartagena, Valle de Escombreras, 30350 Cartagena, Spain

*Corresponding author: f.coletti@hexxcell.com

Refineries turnarounds (T/A) take place at the end of long operating cycles that may last for several years. This extended period of shutdown is used to carry out maintenance activities, revamps, inspection and testing of equipment or debottlenecking projects. Turnarounds are expensive due to the additional cost of labour, materials and contracting and loss of production. As a result, refinery personnel must carry out the planned activities efficiently and often quickly react to unexpected events to allow the plant to return into operation as soon as possible.

Predictive monitoring and prescriptive maintenance are powerful tools that can support decision-making not just during normal operations but also during T/A as they allow predicting the performance upon start-up and detecting inefficiencies of the process once the plant is in operation. In this paper, we present a case study showing the benefits of using such tools during the T/A of a Repsol refinery being monitored with Hexxcell Studioâ„¢.

During the T/A, the predictive capabilities of the monitoring tool were used to predict the temperatures along de train and the distance with respect to the furnace maximum capacity under different scenarios considered for the start-up by operations. This included: a) the possibility of starting up with some of the exchangers bypassed (as they were undergoing repairs); b) the impact that tube-plugging of one of the key exchangers in the network could have on the thermo-hydraulic performance of the network. The results guided the plant engineers to decide on configuration and timing of the start-up considering not only the performance of the train, but also possible effects in downstream processes.

After the start-up, the monitoring system allowed detecting inefficient performance of a heat exchanger that had passed unnoticed with standard monitoring methods. This allowed the engineers to react quickly, inspect the unit and fix the issue, leading to an immediate increase of the heat recovery in the network. The presentation will show the benefits obtained both in energy and economic terms as a result.