Near-wellbore clay fines migration presents a formation damage risk in many gas wells. Fines mobilization can occur due to weakened electrostatic forces on ion exchange with an introduced fluid making them more prone to movement by viscous drag, or where the drag forces are sufficient to physically break or lift clay crystals from their original location, and distribute them throughout the pore network. Fines migration potential is typically assessed via coreflood tests on reservoir core material. Ideally, fines migration tests should be carried out on reservoir core with reservoir gas, at reservoir rate and pressure conditions, but it is often more practical (and less expensive) to scale higher field pressures down to lab scale. However, in order to reproduce the total gas flux observed in a given near-wellbore system, flow rates are routinely increased to counter the pressure decrease. This paper aims to address whether this current lab practice is valid, and to identify alternative yet practical test protocols and fluids that might more closely represent the reservoir gas properties (density and viscosity) that control the drag forces. Reservoir core is often scarce or unavailable, which means that it can be difficult to evaluate different core flood test protocols and fluids. Outcrop samples provide a convenient alternative as they are readily available and cheap to acquire. This paper describes the first phase of a research program that aims to identify outcrop sandstones that are prone to fines migration as a result of drag forces on gas flow, and to evaluate different test protocols. Coreflood tests were carried out on clay rich (predominantly kaolinite) Blaxter sandstone, with samples having a typical permeability of approximately 30-40mD. Potential permeability impairment from fines migration was assessed by sequential and incremental critical velocity tests at both low (290 psig) and high (1450 psig) pressure conditions, and at gas rates of up to 2 L/min. Tests were performed with nitrogen (OFN), and gaseous and supercritical carbon dioxide. In addition, hydrocarbon gas analogues (hexane and dodecane) were also evaluated as a substitute for dense gases in coreflood testing. Initial critical rate tests using KCl brine showed the potential for salinity-related permeability damage in Blaxter sandstone cores, demonstrating that these cores are susceptible to fines migration. However, test results using anhydrous gas demonstrated that pressure and flow rate variation in the laboratory had no notable fines migration effect on the Blaxter sandstone samples. In addition, the use of different hydrocarbon gas analogues showed that even when the test fluid density is selected to so that it is similar to a liquid - supercritical CO2, or light hydrocarbons such as hexane and dodecane - fines migration is still absent even at high flow rates. The outcrop core test results do not necessarily indicate the absence of fines migration potential in gas wells. The kaolinite fines in Blaxter sandstone may not display the well-developed clay crystal structures and morphologies normally associated with reservoir sands, and which may expose the clays to higher drag forces. The case studies presented here will aid in improving coreflood test protocols for assessing formation damage in gas wells. This improved understanding will ultimately enhance the application of core flooding as a tool for identifying formation damage in gas wells.

Core flood tests are regarded as critical to qualification and optimization of scale inhibitors (SIs) deployed in "squeeze" mode, to assess both formation damage and chemical performance. However, the different test protocols commonly adopted can have significant impact on the outcome of both these aspects. Generally, SI core flood tests are designed to obtain both pieces of information from a single flood, often compromising the optimal testing of either or leading to chemical performance aspects being favoured over formation damage orvice versa. Recent reports have illustrated how differences in test protocols can impact chemical performance results for clastic sandstones; the work presented in this current paper examines similar challenges in tight carbonate systems, such as those exhibiting both matrix and fracture flow. It demonstrates the importance of conducting core flood tests under representative conditions for these more reactive substrates in order to qualify chemicals appropriatelysuch that upscaling to the field case can be accurately achieved. A suite of core flood tests were conducted on outcropcarbonate cores under matrix- and fracture-dominated flow conditions (simulating both macro and micro fractures), which allowed examination of chemical behaviour under different application conditions, thus highlighting differences in chemical-retention properties and associated treatment lifetimes as well as in formation damage assessment. This paper examines results from fractured carbonate core tests, which were designed to examine SI interaction and retention where chemical transport is dominated by diffusion, and compares these withsystems where transport is dominated by advective flow in the rock matrix. The overall aim was to examine the impact that core test design can have on the results observed and to discuss the consequences of different test approaches for chemical qualification. In summary, results show that different fracture apertures and flow conditions (matrix versus fracture flow) result in significant differences in formation damage and chemical retention, illustrating the importance of correctly replicating near wellbore conditions when designing such tests.

The requirement for appropriate placement in matrix acidizing to achieve efficient, effective stimulation or removal of formation damage has long been recognized. Despite this, assessing the effectiveness of an acid treatment in terms of placement and treatment efficiency prior to field deployment remains a challenge. This paper presents a scenario where carefully designed core flood tests have been deployed alongside the use of a near-wellbore simulator to model and predict acid stimulation using viscosified fluids for treatments in a range of scenarios. Dual linear core flood tests were conducted to assess placement and clean-up efficacy of a viscosified acid treatment. The efficiency of the treatment was then modelled using a state-of-the-art computer simulator. The computer simulator used is one which has been used extensively to model and optimize scale-inhibitor squeeze treatments in long-reach/complex wells. Its capability was extended to evaluate effects of fluid viscosification (staged or otherwise) on stimulation treatments. Cases where partial/localised clean-up of a damaged zone was achieved, resulting in uneven stimulation were also examined. Results obtained from the model were validated with dual-linear core flood tests using simulated damaged and non-damaged zones. Resulting laboratory injection/diversion was compared to model predictions. Core flood tests highlighted the importance of laboratory simulation of viscosified treatments. These tests showed how improved placement can be achieved by careful fluid design using viscosity alterations. The model was used to demonstrate the benefit (or otherwise) of different degrees of viscosification on evenness of treatment placement for systems with permeability contrasts, pressure differentials, presence of a water-producing zone, and various degrees and depths of skin per zone (including consideration of the effects on placement of acid reactivity on this damage). Results showed that often even modest viscosification of the treatment fluid (~20 cP) improved placement. The greater the fraction that was viscosified (for staged treatments), the more even the placement. Viscosification invariably improved placement where a difference in native permeability existed (as opposed to a difference due to damage) and in most cases reduced the fraction of treatment fluid entering a water-producing zone.

Being able to predict formation damage due to near-wellbore scale deposition relies on accurate scale prediction modelling. It is recognised that the relative paucity of HPHT solubility data can result in inaccurate predictions, as current models typically extrapolate from solubility data obtained on simple systems under more conventional conditions. This paper describes the generation of additional fundamental solubility data under HPHT conditions, and compares the obtained values with several existing models. A purpose-built laboratory test rig capable of making mineral solubility measurements up to 250 °C and 25,000 psi has been used in this work. Experimental methodology has been developed using calcium sulphate to demonstrate that equilibrium conditions have been reached. In this work, barium sulphate solubility data has been generated at conditions up to 200 °C and 20,000 psi in the presence of calcium or magnesium ions, as it was recognised that available data for oilfield-representative brines under such conditions was limited. Scale prediction modelling has been conducted using a range of software widely used in the industry to assess the accuracy of the models in these circumstances. Results for calcium sulphate solubility, obtained earlier, indicated the importance of validating the test methodology, not just for each mineral but also under the required temperature and pressure, to verify that equilibrium solubility conditions have been achieved. Barium sulphate solubility increases with the addition of other divalent cations, but the extent of the increase is at present not accurately predicted by existing scale prediction models. In some cases, the predicted barium sulphate solubility was up to three times greater than the experimentally determined value. It is apparent that there is considerable scope for improvement of scale prediction models under HPHT conditions, particularly in complex brines systems, and that further fundamental solubility data is required to facilitate this. This paper provides additional data for mineral solubility under HPHT conditions with more complex brines, which are more representative of those produced in oilfields. The work demonstrates the limitations of existing scale prediction modelling software under HPHT conditions, particularly in the presence of significant concentrations of other divalent cations, and illustrates areas where additional data and model development are critical to enable more accurate modelling of scaling risk.