US LNG Contract Data: Extract the Option Holder, Not Just the Word Option
Finding an option is only the beginning
A US LNG contract-data workflow should identify who holds an option, what action it permits, and the conditions attached to its use. Detecting the word option is not enough. A summary that assigns the right to the wrong party can reverse the commercial meaning even if it captures the associated dates and quantities correctly.
A filed LNG bunker-vessel charter separately identifies a charterer's extension option and an owner's heel-retention choice. The parties do not hold every option mentioned in the same document. SEC-filed charter
That example is specific to one agreement and is not a rule for other LNG contracts. It demonstrates why a structured summary needs the actor as well as the action. The commercial team should be able to distinguish its own rights from choices another party may exercise against it.
Represent the whole decision structure
A useful option record answers several connected questions: who may act, what they may do, when they must act, what conditions apply, and what changes if they act validly. The source passage and relevant definitions should remain accessible. The goal is to help a responsible reviewer assess the provision, not to replace the contract with a simplified database entry.
For a delivery option, the action might concern quantity, destination, or timing. Those dimensions should remain separate. A right to change quantity does not imply a right to change the loading window, and permission to request another destination does not necessarily compel the other party to accept it.
The wording surrounding the action can change its meaning. A party may request, elect, propose, or agree to a change. An extraction system should preserve that language and flag uncertainty rather than normalize every expression to a single affirmative right. The final interpretation belongs with the responsible contractual advisers.
A fictional comparison of three provisions
Consider three invented summaries for a US LNG cargo agreement. In the first, the buyer may elect an additional quantity by a stated deadline. In the second, the buyer may request an additional quantity, subject to the seller's agreement. In the third, the seller may offer an additional quantity that the buyer can accept or decline.
All three concern additional cargo, but they create different commercial planning positions. The first may support a buyer-controlled quantity scenario if the actual conditions are met. The second depends on a later agreement. The third depends on the seller first making the offer. A database that marks all three as buyer quantity option loses the distinction that matters.
Assume the contemplated increment is a fictional 400,000 MMBtu. A downstream customer offers an additional contribution of $800,000 if that quantity can be supplied, before an invented $250,000 of incremental shipping expense. The potential net contribution is $550,000. That arithmetic does not establish that the increment is available under any of the three provisions.
The trader should therefore keep the $550,000 as a conditional scenario until the relevant right or agreement supports it. If the model treats the amount as secured merely because an option-related phrase was found, it converts a possible commercial benefit into an unsupported forecast. The extraction error becomes an economic error.
Distinguish existence, exercise, and effect
The existence of an option does not mean it has been exercised. A message discussing whether to use it is not necessarily a valid notice, and a draft notice is not evidence that it was sent by the authorized party. The workflow should preserve these states rather than moving directly from clause found to option exercised.
The consequences also need their own treatment. Exercising one right may change the price, quantity, period, or another obligation. The summary should link the effect to the action and retain any cross-reference needed to understand it. A deadline without the associated consequence can be misleading, just as a price without its trigger can be misleading.
Where a provision depends on a defined event, do not substitute an assumed calendar date unless the evidence establishes that event. A date calculator can assist after the relevant inputs and counting method are confirmed. It should not resolve contractual ambiguity by selecting the interpretation that produces the cleanest schedule.
Give human review a specific job
NIST's AI Risk Management Framework calls for defined processes for human oversight and documentation of system knowledge limits. A review step needs an explicit purpose and responsible role. NIST AI RMF
For LNG option extraction, the reviewer should verify the party, action, conditions, timing, and commercial effect against the relevant text. This is more effective than asking someone to approve a polished paragraph without showing its evidence. Material uncertainty should remain visible until resolved, even if the software can produce a confident-sounding answer.
Testing should include provisions where similar words lead to different rights. It should also include amendments that change one element while leaving others intact. A useful evaluation measures whether the system preserves the contract's decision structure, not merely whether its summary resembles a familiar clause.
The workflow should never send an exercise notice or commit a commercial change merely because it has extracted a potential right. Those are separate actions requiring the appropriate authority and review. Keeping extraction and action distinct makes the system more useful by allowing teams to inspect the result before a consequential step.
Zeaclub's ambition for intelligent maritime workflows can be grounded in this practical standard: make contract information easier to locate and compare without inventing rights. This describes a design direction rather than a deployed legal-review capability. Bring Zeaclub a representative LNG workflow and the decisions your team needs to support so the conversation centers on reliable commercial assistance and clear human responsibility.
Sources
- SEC, Exhibit 10.16, LNG bunker-vessel charter. Different option holders in one agreement only; checked October 23, 2026.
- NIST AI RMF 1.0. Human oversight and knowledge-limit context only; checked October 23, 2026.
Related articles
- 0469: US LNG Enquiry Automation: Keep Conditional Cargoes Out of the Firm Supply Column
- 0461: US LNG Atlantic Options: Value the Destination You Can Actually Deliver To
- 0431: US Tanker Enquiry Extraction: Preserve the Quantity Option Before Searching


