(635h) Financially Risk-Aware Plant Maintenance Turnaround Planning Incorporating Reliability in Integrated Chemical Sites

Maintenance turnarounds in processing plants ensure reliable operation and improve plant productivity. A turnaround may include large sets of preventive maintenance tasks, routine component replacements and additions, capacity expansions as well as discovery work and corrective actions. As a result, site-wide disruption from planned shutdowns are common. In practice, turnaround projects are managed by dedicated steering committees, possibly with additional help from third-party turnaround experts in planning and execution, but often with suboptimal coordination across the network. However, in a large chemical site network, planning and coordinating turnarounds for the various units from a high-level standpoint has potential benefits [1,2].

In our recent works [3,4], the focus has been quantifying the risk of loss in rescheduling a turnaround, which offers flexibility in strategic medium-term planning of turnarounds. Here, uncertainties in plant or unit reliability in the form of unplanned outages are considered. In [3], we presented a reactive or a myopic planning strategy and an anticipative planning strategy to quantify the risk of loss in an alternative turnaround schedule instead of a prior base schedule. The multistage stochastic linear programming (MSP) based anticipative model was shown to capture production planning better in the decision-making process via cumulative profit probability distribution profiles. In [4], we compared a portfolio of risk-functionals for the MSP objective on simple case studies along with a few sensitivity study for the timing of the turnaround reschedule.

In this work, we extend the stochastic programming model in [3,4] to a mixed-integer linear programming (MIP) model for medium-term turnaround planning in integrated sites that also takes into account plant reliability. The timing of the turnaround determines the unplanned outage scenarios, and thus, gives rise to endogenous uncertainty. We enforce equivalence of appropriate scenarios via logic constraints, which are analogous to nonanticipativity constraints, using the binary turnaround decision variables. The new formulation also considers multiple reliability-driven turnaround units. We consider different risk-averse objectives in the new anticipative model, and compare profit distributions for the optimal schedule with the myopic risk evaluation strategy. We also consider a simulation strategy to generate a profit profile that is synonymous with value of perfect information in two-stage stochastic programming within the current setting.

References

[1] S. Amaran, N. V. Sahinidis, B. Sharda, M. Morrison, S. J. Bury, S. Miller, and J. M. Wassick. Long-term turnaround planning for integrated chemical sites. Computers & Chemical Engineering 72, 145-158, 2015.

[2] S. Amaran, T. Zhang, N. V. Sahinidis, B. Sharda, S. J. Bury. Medium-term maintenance turnaround planning under uncertainty for integrated chemical sites. Computers & Chemical Engineering, 84, 422-433, 2016.

[3] S. Rajagopalan, N. V. Sahinidis, B. Sharda, S. Amaran, and S. J. Bury. Flexible turnaround planning in integrated chemical site networks. 2015 AIChE Annual Meeting, Salt Lake City.

[4] S. Rajagopalan, S. Amaran, A. Agarwal, N. V. Sahinidis, B. Sharda, S. J. Bury and J. M. Wassick. Financially Risk-Aware Strategies for Rescheduling Plant Maintenance Turnarounds in Integrated Sites. 2016 AIChE Spring Meeting and 12 th Global Congress on Process Safety, Houston.