High-Value Data at Zero Cost: Pulse and Interference Testing in a Deepwater Gas Field Under LNG Plant Start-Up Constraints

摘要

Commissioning and start-up (CSU) of an all-subsea development of offshore gas reserves feeding into an onshore liquefied natural gas (LNG) plant is complex and challenging - yet project success is driven by a timely and flawless LNG project start-up. To this end, development wells in the Jansz-Io gas field had to be flowed back to the onshore facility to prove up Gorgon LNG Train 1 capacity before the wells could be relied upon for uninterrupted gas supply to Train 1 start-up. Here, a carefully-considered well proving sequence can yield unique, once-in-a-lifetime and zero-cost subsurface data that improves reservoir characterisation and future decision quality. This paper describes: 1. Design aspects of the well proving sequence that enabled opportunistic pulse and interference tests to be conducted at no expense to Train 1 start-up: 1.1 Optimality: balancing the well proving objectives of the various functional stakeholders (flow assurance, facilities, operations and subsurface) while honouring all CSU / operational constraints. 1.2 Flexibility: recognising and planning for CSU uncertainties (subsea gathering system versus onshore plant readiness; onshore gas demand; and pigging schedule for liquids management). 1.3 Contingencies: planning for sequence execution issues such as well unavailability and gauge failure(s). 2. Execution - how the tests were executed and how the reservoir responded. 3. Analysis and interpretation of pulse and interference test data, whereby: 3.1 A simple, conceptual analytical model based on the line source solution was constructed and used to guide analyses in more detailed numerical models. 3.2 A single-layer 2D numerical reservoir model with local grid refinement at the wells was constructed and used to infer inter-well average reservoir storativity (porosity-total compressibility-thickness product, ?cth) and transmissivity (mobility-thickness product, kh/μ) and across-fault communication. 3.3 Existing 3D models were calibrated to the pulse and interference testing data for improved reservoir performance predictions in support of future reservoir management decisions.