The Center of Expertise in Economic Modeling and Studies (CEEMS) of GDF SUEZ has used the AIMMS modeling system to develop a management tool for an LNG terminal, ensuring fair treatment of the multiple users of the terminal, while maximizing production and flexibility.
Production planning & flexibility allocation at multi-user LNG terminal
Natural Gas can be transported in liquid form (LNG) by ship, which enables gas companies to work with a diversified portfolio of supply sources. At marine terminals, the LNG transported by ship is unloaded into cold storage tanks and then regasified and delivered to customers via a gas distribution network. GDF SUEZ and Total have jointly invested in such an LNG terminal with a capacity of 8.25 Gm3/year at Fos-Cavaou in France, handling more than 100 ships per year. The terminal is managed by STMFC (French acronym for Society of the Fos-Cavaou Gas Terminal) and is currently used by several gas companies, which can all unload, store and regasify LNG at the terminal.
Managing the multi-user LNG terminal presents a complex challenge: physical constraints of the terminal must be respected, but also commercial conditions of the contracts with the terminal’s users.
- Physical constraints include limits on the allowed gas production levels, smooth production level variations, and minimum & maximum storage quantities.
- Commercial conditions include a fair treatment of all users in planning production, allocating the effects of unforeseen changes, and offering flexibility.
Designing a plan for the terminal that meets all physical constraints and commercial conditions is impossible to do by hand. Moreover, plans need to be adjusted frequently, so an automated decision-support tool is needed.
The CEEMS team has used the AIMMS modeling system to develop a planning tool called “Opti-ATM” (“Accès au Terminal Méthanier”, i.e. Gas Terminal Access) that optimizes the planning of the LNG terminal in three stages:
- Quarterly production planning. For a given schedule of LNG ship arrivals, the daily production for each user is planned three months ahead. The chosen approach for the Opti-ATM tool consists of two steps: first the overall production level is optimized; then the production is partitioned among the users of the terminal.
- Planning revisions. The plan needs to be revised often as a result of technical difficulties or changes in the LNG ship arrivals of one of the users. The Opti-ATM tool uses two MIP models: first the impact on the other users is minimized; then the impact on the user at fault (if it exists) is minimized.
- Calculation of flexibility. The terminal must communicate how much up- and downwards flexibility each user can use without disturbing the planning of other users or the overall planning. The Opti-ATM tool performs three stages: the daily overall flexibility, the daily flexibility per user, and the cumulative flexibility per user.
The Opti-ATM tool automatically proposes plans for the LNG terminal that meet all physical and contractual constraints. These plans lead to higher production levels and higher flexibility without incurring contractual penalties for unfairness, which could not have been achieved without the tool. In addition, the flaring of excess gas is reduced and the operation of the terminal is smoother, so the lifetime of the equipment is prolonged.
How did AIMMS add Value?
AIMMS enabled the CEEMS to develop quickly a prototype tool that was close to the professionals in the field, as it includes the physical and contractual constraints. Various mathematical modeling approaches (MIQP, MIP, LP, heuristic algorithms) could easily be tested within the same modeling environment before the final modeling approach was chosen for the planning tool based on the best results from an operational perspective. The final planning tool was deployed to the end-users with the integrated Graphical User Interface developed in AIMMS, enabling them to see the planning results immediately in a graphical form.
Anna Robert, model developer in the CEEMS of GDF SUEZ, said: “We used AIMMS to design the complete model of the STMFC problem. It includes several steps, therefore, several models have to be solved one after the other. AIMMS is easy to use for changing the sets of constraints, trying new objective functions. It permitted to reach the model which is now used by operational teams.” Benoît Labaune, manager of STMFC, said: “Using AIMMS, and specifically its integrated GUI, helped us in our comprehension and analysis of the optimization model developed by the CEEMS. During the model design phase, we have been able to explain easily what was fine in the model and what has to be modified. The discussion with the developers was facilitated by the flexibility of AIMMS.”
GDF SUEZ is one of the leading energy providers in the world and active across the entire energy value chain: in electricity and natural gas, upstream to downstream. It has 200,000 employees throughout the world. The team in charge of the Opti-ATM is part of the CEEMS, which is composed of about 100 people. The CEEMS is based both in St Denis (France) and Brussels (Belgium). It develops and delivers decision support tools for various business departments of GDF SUEZ and it also provides strategic studies for the group.