Hydrocarbon reservoirs are typically porous rocks where oil and gas are filling the pores, like a sponge. To produce the hydrocarbons, operators drill wells to penetrate the reservoirs much like water wells are used to produce water from aquifers. However, the reservoir rocks lie often at depths between 2 and 5 km. At such depths, the wells and the rocks are exposed to enormous stresses due mainly to the weight of the layers above them. If these stresses are too high compared to the strength of the rock around the well, the rock will collapse. As a result, sand grains and particles from the sandstone will detach from the rock and will flow out of the well together with the produced oil or gas. This is called sand production and it may create serious operational problems such as well loss, erosion of pipelines and process facilities, as well as environmental and safety problems. A common solution is to install sand control equipment like filters or screens. However, allowing a controlled rate of sand production is possibly more cost effective due to lower capital costs and because it enables higher production rates.
The project studies sand production using a novel laboratory device that mimics the reservoir conditions at 2 to 5 km depth in terms of the high and anisotropic stresses and the in-situ reservoir pressure. The laboratory device is unique in the world and has been specially designed by SINTEF for this type of studies. The produced experimental results have helped us prove our hypothesis that in the field sand occurs earlier in the fields life but with less amount of sand than what earlier classical sand studies had shown. These help us improve the prediction of the onset of sand production and, most importantly, the sand rates in the field. The ultimate goal of the project is to avoid or manage sand production in a way that operators can improve production rates, and hydrocarbon recovery and at the same time reduce operating costs, reduce the environmental footprint and reduce drilling costs by extending the life and productivity of existing wells.
The Project kick-off Meeting took place on 01 Nov 2017. A PhD student started in December 2017 and will defend his PhD Thesis in early 2022. New industrial partners joined the Project on Fall 2017 and 2019 bringing the total to five industrial partners, three universities and SINTEF. The project has three Steering Committee meetings per year with one of them a physical meeting and the others videoconference meetings. The physical meetings have been two-day events. In 2018, a workshop was held in the second day where the industrial partners presented their problems and showed how the project contributes to improve their competence and operations. The workshop ended with a visit to the lab where the new True Triaxial sand test facility was presented together with a live sand production test. In 2019, a workshop was organized where the participants had hands on instruction on the use of the finite element software Geo3D developed in the project. We also had working meetings with our partners in Trondheim, Houston and Tokyo and participated with papers and abstracts/presentations in several conferences. The Final meeting in fall 2021 took place online due to travel restrictions.
The project results have shown the effect of stress anisotropy on sand onset and mass and confirmed our hypothesis that stress anisotropy results in earlier sand but less sand. The results were analysed and used to calibrate our semi-analytical, log-based SandPredictor software and our finite element numerical models. They can explain the observed discrepancies between laboratory and field calibrated models. Results using various saturating and flowing fluids have shown the effects of watercut on sand onset and sand production and how this can be incorporated in our predictions. Cavity pack tests have shown the importance of the proppant pack inside the perforation and how both poor and good proppant packing may lead to productivity loss. SINTEF's SandPredictor software has been updated and to a modern, robust, and user-friendly software and new features are continuously added. Finite element method software has been developed and tested by the partners.
A new generation semi-analytical, log-based, user-friendly SandPredictor software has been developed. The software has large impact on the industrial partners as it is adopted as their preferred day-to-day sand analysis tool. Two numerical finite-element numerical models have also been developed.
The hypothesis that anisotropic stresses give earlier sand onset in the life of a well but with less sand production was confirmed experimentally. This outcome has a profound impact in the design of well completions, as has the effect of fluid saturation and fluid flow and especially watercut.
The capability of the state-of-the-art True Triaxial Apparatus at SINTEF in reproducing successfully the in-situ stress and flow conditions in deep reservoirs was proven. International research collaboration was promoted involving scientists from Norway, USA, and Japan. Several MSc and three PhD students were involved in the project gaining valuable experience and expertise.
Sand production has been an area of research in the past 25 years due to its significance in the production of hydrocarbons in sandstone reservoirs. Operators are continuously pushing the boundaries to improve the production and recovery rates and reduce the operating costs by increasing the applied drawdown, avoiding or limiting sand control and applying sand management methodology to manage sand production in connection with improved techniques for separating the produced sand. Moreover, depletion in mature fields, such as in Norway, exasperates the problem. Increasing profitability and recovery rate from such fields is a target of the project. An improved sand prediction technology based on next generation analytical and numerical software can lead to considerable commercial benefits from increased productivity and recovery and reduced up front capital expenditures.
Research in sand production has addressed various issues. Despite the progress, sand models have not been proven under fully anisotropic stresses which are commonly encountered in the field. In fact, the majority of models have been experimentally validated under isotropic stresses and subsequently applied to anisotropic stress states. Moreover, in frac-pack completions no real tool exists to analyze and predict sand.
The activities to reach the objectives are based on an experimental part, which will establish sand onset and rate in conditions as close as possible to the field. Given that field sand production data are unreliable, the proposed experimental program will be the next best case to reality. Frac-pack completions will also be simulated in the lab. The lab results will be used to develop, validate and qualify next generation sand prediction analytical and numerical models and tools, such as SandPredictor and Geo3D, and enhance them with frac-pack modeling capability. Both models are well established in the industry and they offer the state of the art in sand prediction.