Productivity loss of gravel-packed wells have been reported for many cases in the North Sea and worldwide. The problem is here studied through physical model experiments of gravelpacked perforation cavities in weak field - and outcrop sandstones with emphasis on the interaction between the gravel-pack and the formation rock. The gravel filling efficiency has been varied between 0 and 100%. The flow productivity of the simulated formation and of the gravel itself has been monitored during tests at realistic in situ stress-fields, pressure gradients, and fluid flowrates. The relative effects of in situ geometry, water breakthrough, rock failure and sand release mechanisms, and gravel contamination have been investigated.
The experiments have revealed the effect of intrinsic rock failure and sand production mechanisms on both formation and gravel-pack productivity. Contamination of the gravel with formation sand and fine particles appeared not to dominate the productivity variations, which mainly were affected by effective stress changes through variation in drawdown and depletion. The experimental results indicate that imposed restrictions on rock dilation due to high gravel filling efficiency, therefore restrict the fluid flow more than do the contamination of the gravel by sand and fines. Thus the overall productivity should be considered and not only local flow barriers such as contamination of the gravel. The emphasis in modeling must accordingly be given to sand release mechanisms and inherent formation permeability changes due to rock dilation and creation of cavities between the gravel and the formation, as well as internal plugging by fine particles, relative permeability effects, etc.
The observed effect of the gravel filling efficiency have thus to some extent led to results diverging from the classical approach aiming at maximization of the gravel filling efficiency while sizing the gravel according to some filtration rule. As a result, the use of gravel as a filter may lead to different design criteria as compared to the case of borehole re-enforcement (100 % filling efficiency).
In addition to the various gravel packing techniques and tools, a large literature exists governing the issue of gravel pack optimization by matching of the gravel size to the formation grain size (see e.g. Saucier, 1974, Suman and Snyder, 1982, Stein, 1983) and optimizing the gravel filling efficiency, i.e. the degree of filling of the perforation cavities (for perforated liner completion) with gravel. In spite of field observations of productivity loss interpreted as due to fine particles blocking the gravel (Unneland and Waage, 1991), the design criteria for gravel sizing are still primarily based on the ratio of the mean gravel size to e.g. the mean or the median grain size of the formation or, alternatively, any percentile of the finest sand to be screened.
Previous studies (Saucier, 1974, Penperthy and Cope, 1980) neglect to some extent two important circumstances that may affect the productivity of the wellbore significantly:
Permeability alterations in the formation surrounding the gravel due to changes in stress geometry due to drawdown and depletion and due to particle release upon sand production.
Released particles may not reflect the virgin grain size distribution of the formation due to perforation induced grain crushing, transport and deposition of micro-particles in the near-wellbore region and creation of larger rock fragments as determined by the rock failure mechanism.
This means that the gravel pack design rule should be modified depending on other formation characteristics than purely the virgin grain size distribution as described through a sieve analysis.