Flexible turnaround planning for integrated chemical sites

Transcription

1 Flexible turnaround planning for integrated chemical sites Amaran et al., 25 Sreekanth Rajagopalan, Nick Sahinidis, Satyajith Amaran, Anshul Agarwal, Scott Bury, John Wassick Enterprise-Wide Optimization (EWO) Meeting, CAPD 26 March 8-9, 26

2 Turnaround rescheduling Motivation Respond to peak demands High forecast during turnaround period Unfavorable market conditions Foreseeable supply/demand variations Additional resource constraints Availability of skilled workforce or process specific technical experts Problem statement Given an integrated chemical sites network with a base turnaround schedule over the next 6-9 months, Benefit of moving Unit 4 turnaround from March to July? Risk of loss in rescheduling? Unit 4 Unit 2 S Unit 9 Continue operations Performance exceeding expectations (less catalyst activity degradation, HEX fouling) S2 S3 Unit 5 Unit 6 Unit 7 Upstream Downstream Storage Tanks 2

3 % Cases (N=25) Additional days worth production Potential benefits Key factors Time value of money on costs and revenue Time to plan production and inventory better Integration effect Unit 9 Unit 7 Unit 6 Unit 5 Base Unit Unit Jan Feb Mar Apr May Jun Jul Aug Sep Unit 4 turnaround in March Alternative 2 5+% 55+% 6+% 65+% 7+% 75+% 8+% 85+% 9+% 95+% Demand Range 2 Unit 9 Unit 7 Unit 6 Unit 5 Unit 4 % cases Alternative profitable % cases Alternative profitable (4+ (>=%) days) Unit 2 Avg improvement Avg - S.D. Improvment Jan Feb Mar Apr May Jun Jul Aug Sep Avg + S.D. improvement Unit 4 turnaround in July 3

4 Probability Planning under uncertainty: unplanned outages.3.2. Unplanned Outage Severity March April May June July No outage Pit stop 6 No outage Pit stop Full TA Corrective maintenance decision policy: pit stop (minor) and turnaround (major) Reactive planning vs. anticipative planning t = t = Pit stop (CM) Turnaround (CM) Turnaround (PM) No outage Full TA 5. t = 2 Pit stop 4.3 t = 3 No outage Full TA 3. t = 4 Pit stop Full TA 2.9. Full TA.49 9 t = Mar Apr May Jun Jul Scenarios 4

5 Stochastic programming model max NPV Flow balance Demand constraints : expected value of all scenario profits : material balance + stream ratio requirements : upper bound (deterministic, monthly timescale) Example: nodes, 6 arcs, 3() turnarounds, 9 month horizon 72 time periods 33 stages (4-month reschedule window) 2 time periods per stage Turnaround constraints : unit up or down for maintenance Capacity constraints : flow and storage tank bounds Nonanticipativity constraints : time-consistency and implementable decisions Model tractable to reliably reschedule single turnaround: stages ~ 2T Solution time for deterministic equivalent of SP is < 2 sec Base Alternative Reactive (sequential LPs) Anticipative (multistage SP) variables,48,48 27,648 54,736 constraints,987,987 38, ,35 non-zeros 5,44 5,44 5,28 96,53 5

6 % Cases (N=25) Additional days worth production Probability of NPV <= X Potential reschedulable instances NPV improvement: Anticipative vs Base Risk profiles: cumulative distribution of profits % 55+% 6+% 65+% 7+% 75+% 8+% 85+% 9+% 95+% Demand Range % cases Alternative planning profitable % cases Anticipative planning profitable % cases Anticipative planning pl. profitable profitable (>=%) (4+ days) Avg improvement Base 299 Alternative 38 Reactive 34 Anticipative Reactive planning NPV (scaled cost units) Anticipative planning -3% instances can be reliably rescheduled More than 4 days worth production recovery in about 5% cases Anticipative planning model provides a flexible production and inventory plan that is less risky 5% chance of loss vs. % from reactive planning 6

7 Tank 2 inventory level (%) Tank 3 inventory level (%) Additional inventory to hedge against uncertainties Inventory with time for tanks 2 and 3 for the (final) scenario corresponding to no outages Tank 2 Tank Jan Feb Mar Apr May Jun Jul Aug Sep Jan Feb Mar Apr May Jun Jul Aug Sep Time Time Anticipative Reactive Base Anticipative Reactive Base Anticipative plan recommends more inventory to hedge against future uncertainties Small premium of.5 scaled cost units incurred from additional holding cost 7

8 Probability of NPV <= X Sensitivity to outage probabilities What if outage probabilities data is underestimated? Anticipative NPV (scaled cost units) Reactive NPV (scaled cost units) [ 7 different profiles. Dark to light :: best- to worst- case] Change in probability of loss compared to base schedule is within 5% Anticipative planning model values reschedule more than best-case reactive plan 8

9 Probability of NPV <= X Sensitivity to reschedule time window Effect of reschedule time ( to 4 months) for decreasing demands NPV (scaled cost units) 3-month window (5%) AP-4 AP-3 AP-2 AP- RP-4 RP-3 RP-2 RP- Short window (-2 months) risky (25-35% chance of a loss) when demands not sufficiently low Long window (4 month) risky (5% chance of a loss) due to corrective maintenance costs 9

10 Conclusions Rescheduling a turnaround offers production recovery as high as -2 days Depends on demands as well as integration effects Anticipative planning model hedges against uncertainties due to outages at a small premium Timing of reschedule as well as performance condition of the unit affects potential cost benefits and risk of loss Future work Optimal turnaround reschedule time window Simultaneous condition-based and risk-based turnaround planning Practical-scale networks