Abstract

Surface movement data can be used in data assimilation or inversion exercises to improve the level of knowledge of a compacting reservoir. We have designed, implemented and tested a new algorithm that uses measured optical height differences directly, without having to translate them to heights with regard to a geodetic reference level. This way, the measurement points are independent of reference benchmarks. In addition, the accumulation of inaccuracy of the height difference measurements, which have to be connected to a stable benchmark over multiple connected benchmark pairs, is circumvented. Thus, the new method is less error prone. Individual errors in height difference measurements do not propagate into interpreted heights. The procedure was applied to a synthetic case study and compared with the traditional inversion using interpreted height measurements. The reduction in standard deviation between prior and estimated values for the different adjusted parameters was better for the new procedure.

1. INTRODUCTION

The production of gas from subsurface reservoirs in many cases induces ground movement. The pressure reduction results in reservoir compaction, which is transferred to the surface and visible as subsidence. Ground movement is traditionally measured with optical levelling technology. Well-founded benchmarks are installed, and in repeated measurement campaigns their respective positions are measured. The heights of the benchmarks visited at each campaign are determined by relating them to a single, stable benchmark with a known height with respect to a geodetic reference system like NAP (the Dutch reference system). Surface movement is then estimated by subtracting heights that have been estimated in measurement campaigns at different times.

Subsidence estimates are important for the assessment of the damage caused by, for instance, changing relative groundwater levels and related intrusion of salt water in areas close to the sea. They can, however, also be used to improve our understanding of the physics of the subsurface. This is important for improving subsidence predictions resulting from ongoing operations, but also for reservoir management. Recent studies have demonstrated the feasibility of using subsidence estimates for the quantification of aquifer activity and fault sealing properties [1, 2].

This content is only available via PDF.
You can access this article if you purchase or spend a download.