Terminals are mission critical assets for effecting mass transfers, exploiting arbitrage opportunities, blending, mopping up errors and inaccuracies in supply chain planning, amongst others, across the hydrocarbon processing value chain. Effective scheduling of terminal operations is required to handle timing, sizing, allocation and sequencing decisions involved in connecting the "ideal" (production plan) to the "real-world" (the operation), with its various subtleties, nuances and non-linearities. Complex decision-making is required to make money. It involves the development of a detailed (executable) plan that is able to implement the operations strategy from the planning process, running the asset(s) up against physical and logistical constraints. As a result, schedulers must deal with a large number of inter-related alternatives with high implications in business performance. Wrong moves in the decision chain can set the execution path towards costly disruptions. Only through combinatorial optimization algorithms can this complexity be simplified. This presentation will show how these algorithms can be incorporated into practical business applications and made available to extend the capabilities of scheduling personnel way beyond what can be achieved with current methods. The value captured and how it is achieved will be demonstrated using actual applications in LNG Regasification, Crude Oil Supply and Primary Distribution operations.
Transforming Decision Making in Scheduling of Terminal Operations
1. Enrique Salomone
Supply Chain Scheduling Expert
KBC (A Yokogawa Company)
November 11, 2020
Digital Transformation
of Decision Making in
Scheduling Terminal
Operations
2. Agenda
1. Opportunities for Digital Transformation
2. LNG Regasification Terminal
3. Crude Oil Supply Terminal
4. Primary Distribution
5. Final Remarks
3. Digital Transformation of Value Chain Optimization
Extracted from KBC’s Value Chain Optimization Manifesto
https://www.kbc.global/insights/whitepapers/value-chain-optimization-manifesto/
Focus of this
presentation
Advanced Analytics
Real-time and Astute
Decision Making
Intelligent Sensing
Agile Actions for
Value Creation
4. Opportunities in Terminal Operations Scheduling
■ Commonalities of decision making
◆ Managing inventories with complex logistic operations
◆ Volume and quality variations have strong impact
on business performance
◆ Time chained decisions with explosion of alternatives
■ Digital technology offers clear opportunities for superseding
the traditional way with advanced analytics
5. ■ Sequential decisions: One step at the time
◆ Short sighted
◆ Conditioned by previous decisions
◆ Solving the maze without drone vision
■ Applying ‟known to be good” patterns
◆ Good for business as usual
◆ Poor performance for handling
disruptions or volatile situations
Traditional Way
6. Digitalized Way
Integrated model
of the supply chain
Business
Application
Capturing operations
Decision model
for automated
combinatoric search
Capturing decision’s
space
Capturing business process
9. Calorific Values of Incoming Cargoes
JUL AUG SEP OCT NOV DEC JAN FEB MAR APR JUN JUL
10. Demand at the Different Delivery Points
JUL AUG SEP OCT NOV DECJAN FEB MAR APR JUN
11. Scheduling Decisions
Unload tanks and split
Tank capacity
and berth connections
Inter tank transfers
Tank capacities
and pump connections
Regas injection
Demand, tank
inventory, pump
connections and
quality specs
12. Scheduling Business Objective
LPG
45
44
43
42
41
Terminal
City Gas
Power Plant
Managing the basket of
incoming cargoes with
different calorific value
To satisfy demand
quantity and quality
specifications
Minimizing the need of
additional LPG injection
Through the terminal
logistics and
regasification facilities
13. Decision Model
A high level modelling
language ensures an efficient
translation of business
constraints and alternatives
into a decision model suitable
for automated search
14. Applying Combinatoric Search
■ Case for 31 days, eight (8) cargoes,
18 tanks, seven (7) shipping lines,
eight (8) vaporization lines, three (3)
inter-tank transfer lines
◆ Total Variables 11098
◆ Binary Variables: 5700
■ First feasible solution found after exploring
32256 nodes
■ Iterations with NLP solver
■ Produces 30 feasible solutions after
162,000, optimal in 7-minutes
Actual business application UI
15. Optimized Solutions
■ Automatic creation of a 30-day schedule for all operations in the terminal
■ Granting feasible inventory projections
■ Tracking density and calorific value in every tank and stream
■ Reducing the number of inter-tank transfers and pump setups
■ Reducing energy giveaway
■ Minimizing LPG injection
16. LNG Regasification Terminal
Using combinatorics search
algorithms produce solutions with
8%
LESS CONSUMPTION OF LPG
2 US$/Tn
Business Value
18. Crude Oil Supply Scheduling
LP feed target
Crude segregationCrude receipts
Crude blends
Replenish
feed tanks
Unload tanks and split
Tank capacity and
berth connections
Inter tank transfers
Tank capacities and
pump connections
CDU feed charge
Plan target, tank
inventory, pump
connections and
quality specs
Quality
constraints
Flow
constraints
19. Scheduling Business Objective
Goals
Meet volume target
Feed blend within quality within spec Sulphur content, TAN, metals, etc.
CDU streams within flow and quality limits Hydraulics, contaminants, etc.
Smooth out variations due to logistics
21. Applying Combinatoric Search
■ Case for 30 days, nine (9) vessels,
10 tanks, 10 CDU/VDU Streams +
five (5) swing cuts, 20 properties,
10 hydraulic constraints,
15 components
◆ Total Variables 4681
◆ Binary Variables: 983
■ Iterations with NLP solver
■ Produces optimized solutions
in 4-minutes
Actual business application UI
22. Optimized Results
Optimized:
6 blocks average length 5 days
Current:
8 blocks average length 3.75 days
3875Current Total Charge
3900Optimized Total Charge
23. Example of Optimized Results
Optimized:
6 blocks average length 5 days
Current:
8 blocks average length 3.75 days
24. Example of Optimized Results
Optimized:
6 blocks average length 5 days
Current:
8 blocks average length 3.75 days
25. Business Value
■ Reduction on quality target variability
◆ Cheaper slate of crude-oils being purchased whilst still being able to meet
the quality specifications for all the finished products refined and blended
■ Improved target setting and less disruption on CDU on crude tank switches
◆ Better coordination of unloading, transfers and blend operation produce
longer stable crude runs
■ More accurate composition tracking of crude in tanks meaning more
accurate LP optimization
◆ Enables LP optimization refinement with improved projection of crude
composition along the tankyard
■ Potential linkage to an APC application for handling of crude switching
■ Spot opportunity for crude-oil trades
◆ Additional opportunities to run batches of opportunistic spot crudes
due to better scheduling of the crude-oil blendshop
Potential
savings of
up to 20
cents/bbl
7 MMUSD for a
100 KBD refinery
28. Scheduling Decisions
Model
Vessel
operations
Timing
Inventory
Routing
▪ Loaded/unloaded products and quantity
▪ Compartment assignment
▪ Operation on berth
▪ Overlapping of operations
▪ Arrival/departure dates
▪ Berthing/waiting times
▪ Loading/unloading start/end
▪ Safety stock level
▪ Storage capacity
▪ Terminal visits
▪ Vessel voyages
▪ Available vessels
▪ Current position
▪ Next available routes
29. Applying Combinatoric Search
■ Case for 60 days, six (6)
terminals, 12 inventory points,
five (5) vessels, two (2)
MMBBLS transported
◆ Total Variables: 20199
◆ Binary Variables: 11224
■ Produces optimized solutions
in 3-minutes
Actual business application UI
30. Example of Optimized Results
10%
19%
Baseline Optimized
Terminal Costs MMUSD 1.627 1.445
Transport Costs MMUSD 0.600 0.555
Total Costs MMUSD 2.227 2.000
USD/Transported BBL 1.11 1.02
0.1 US$/bbl
Business Value
60 days
6 terminals
2 MMBBLS transported
32. Final Remarks
• Hard to find good feasible solutions
• Event harder is to be efficiently responsive
against disruptions
Scheduling terminal
operations is challenging!
• Combinatoric search techniques embedded in
business applications can make the difference
Unique opportunities
for digitalization
LNG Regasification
2 USD/Tn
Crude Oil Supply
20 cents/bbl
Primary Distribution
10 cents/bbl
33. The names of corporations, organizations, products and logos herein are either registered trademarks or
trademarks of Yokogawa Electric Corporation and their respective holders.
Digital Transformation
of Decision Making in
Scheduling Terminal
Operations
Thank you