Over the last decade, the challenge of drilling narrow Equivalent Circulating Density (ECD) window wells has put increasing pressure on the performance of the drilling fluid. As a result of this, so-called ‘Fragile Gel’, or ‘Flat Rheology’ fluids have been developed by the industry and have become widely utilised across the globe. Although the design of these fluids is primarily aimed at optimising performance in deepwater environments, the lower pressure fluctuations imposed on the formation (as compared to more conventional invert emulsion fluids) that is attributed to these fluids has contributed to their increased use in other, non-deepwater areas, where ECD management is still viewed as critical. Although the perceived benefits are now widely understood and accepted, the industry has not yet developed a comprehensive and universally-accepted benchmark for the properties that satisfactorily define these fluids. The authors of this paper believe this to be due to a number of factors, including the different approaches taken within the service industry in developing chemistries that either significantly reduce, or even completely replace, the requirement for conventional organophilic clay-based viscosifiers. BP have developed an internal set of guidelines based on operational experience in regions where these fluids have been successfully implemented; however, these have limitations when universally applied, particularly as new fluids featuring novel chemistries are introduced. This paper discusses the key engineering parameters used to define the engineering guidelines for what BP refers to as ‘Flat Rheology Fluids’. A description of these parameters and how data collected from recent field trials in Norway is being used to help refine and validate these parameters is also described.

In this paper we describe a simulation model for computing the formation damage imposed on the formation during over-balanced drilling. The main parts modelled are filter cake build-up under both static and dynamic conditions, fluid loss to the formation, transport of solids and polymers inside the formation including effects of pore lining retention and pore throat plugging, and salinity effects on fines stability and clay swelling. The developed model can handle multi-component water-based mud systems at both the core scale (linear model) and the field scale (2D radial model). Among the computed results are fluid-loss versus time, internal damage distribution and productivity calculations for both the entire well and individual sections. The simulation model works in part independent on fluid loss experiments, e.g., we do not use fluid leakoff coefficients, but instead we compute the filter cake buildup and its flow resistance from properties ascribed to the individual components in the mud. Some of these properties can be measured directly, such as particle size distribution of solids, effect of polymers on fluid viscosity and formation permeability and porosity. Other properties, which must be determined by tuning the results of the numerical model against fluid loss experiments, are still assumed to be rather case independent, and once determined they can be used in simulations at altered conditions as well as with different mud formulations. A detailed description of the filter cake model is given in the paper. We present simulations of several static and dynamic fluid loss experiments. The particle transport model is used to simulate a dilute particle injection experiment taken from the literature. Finally, we demonstrate the model's applicability at the field scale and present computational results from an actual well drilled in the North Sea. These results are analysed and it is concluded that the potential impact of the mechanistic modelling approach used is (a) increased understanding of damage mechanisms, (b) improved design of experiments used in the selection process and (c) better predictions at the well scale. This allows for a more efficient and more realistic pre-screening of drilling fluids than traditional core plug testing. Introduction A simulation tool, referred to as Maximize, has been developed for the purpose of investigating fluid loss to the formation during over-balanced drilling and the impairment imposed on the formation by the invading fluid. The objective of the program is to serve as a tool for supporting well planners' decisions related to the choice of well fluids, and integrating, analyzing and interpreting laboratory as well as field formation damage data. The filtration properties of the mudcake forming at the wellbore surface has been investigated by several authors over the years, see for example Ferguson and Klotz (1954); Outmans (1963), Bezemer and Havenaar (1966), Arthur and Peden (1988), Fordham et al. 1991 and Dewan and Chenevert (2001). The common understanding is that once the filter cake has been formed, it will control the filtration rate independent of the formation properties, except at very low permeability where the flow resistance offered by the formation is comparable to the filter cake resistance. The filter cake properties depend only on its composition, the pressure drop over the cake Dp and the shear stress acting on the cake surface by the circulating mud. Under dynamic conditions, the filtration rate will approach asymptotically a limiting steady-state rate which only depends on the shear stress at the cake's surface. A combined filter cake and simulation model was used by Semmelbeck et al. (1995) for computing the fluid invasion profile along the well. Further improvement of the filter cake model was presented by Dewan and Chenevert (2001), who demonstrated the derived model's capability to reproduce complex laboratory experiments with sequential changes in dynamic shear rate and overbalance pressure. Others, have also investigated filtrate invasion by numerical simulations e.g., Ding et al. (2002), Wu et al. (2004) and Suryanarayana et al. (2005).

TX 75083-3836 U.S.A., fax 01-972-952-9435. AbstractDrilling the 12 1/4" sections for the Kharyaga oilfield in the Timan-Pechora region of Russia has always been considerably complicated by wellbore instability and hole cleaning issues. These issues have been most serious when drilling through Triassic and Permian shales, sandstones, argillites and silts, followed by drilling Carboniferous limestones and dolomites. Unstable formations sloughing and packing off, wash outs and severe caving has resulted in many problems including drill pipe pack-offs, high torque and drag, the need for additional reaming operations, and difficulties in reaching bottom with casing.Fluid treatments of swelling shales with chemical inhibitors helped considerably with shales hydration, but the problem of caving and packing off still persisted. High angles (up to 52.8°) and long section lengths (up to 2,707 m) aggravated the problems. The key remedy was determined to be effective hole cleaning. Cleaning efficiency of different types of sweeps was studied during drilling of 10 Kharyaga wells. Sweeps were pumped on a regular basis in drilling intervals of 100, 150, 200 and 300 meters, prior to pulling out of hole and when indications of packing off had been observed.Pumped cleaning sweeps included high-viscosity or highdensity single sweeps; tandem sweeps (low-viscosity following by high-viscosity or high-density sweeps); and sweeps with special additives (carbon-based LCM material or innovative monofilament fiber sweeping agent). Also, special attention was given to combined tandem sweeps, which are low-viscosity sweeps treated with a monofilament fiber sweeping agent followed by a high-density sweep (sometimes treated with carbon-based LCM material).Investigation of different sweeps performance showed that the best hole cleaning results for Kharyaga field wells were achieved by circulating combined tandem sweeps, which are low-viscosity sweeps treated with a monofilament fiber sweeping agent followed by a treated or untreated high-density sweep.

BP in the U.K. North Sea attempted to drill a horizontal oil producer in the Harding field in 2006. Designed to produce an initial 10,000 blpd from 2000 feet of reservoir section before water break through, the well was compromised by the collapse of the lower hole due to chronic shale instability. This also resulted in the pre-drilled liner being stuck and set higher than initially planned. With less than 200 feet of reservoir sand exposed at the heel of the well (in close proximity to the water leg), the initial expectations were of 4,000 blpd. However the well was found to be badly impaired and produced 400 blpd. Although the well had been displaced to a carbonate based low solids oil based mud (LSOBM) prior to completion, a significant quantity of the barite weighted drilling system was still in the well. The damaging mechanism was determined to be synthetic oil based mud compressed around the screen completion as well as the mud from the uncompleted horizontal lower hole being squeezed into the screens as the open hole gradually collapsed with time. A Coiled tubing (CT) intervention was carried out in late 2006 with solvents and multiple attempts with an acidic nano wash solvent system, this was not successful in restoring well productivity. In 2007 BP chose to use an advanced chelate based barite/carbonate dissolver system behind a proprietary pre-flush system in an attempt to recover well productivity. Four operations have now been performed since September 2007 without CT, all as simple bull head operations. As a result of these treatments the well productivity (PI) has increased from 1.5 up to 12 blpd/psi. Current well rates are between 4000 to 6000 blpd depending on well stability and slugging caused by increasing water cuts.

In this paper, we describe a simulation model for computing the damage imposed on the formation during overbalanced drilling. The main parts modeled are filter-cake buildup under both static and dynamic conditions; fluid loss to the formation; transport of solids and polymers inside the formation, including effects of porelining retention and pore-throat plugging; and salinity effects on fines stability and clay swelling. The developed model can handle multicomponent water-based-mud systems at both the core scale (linear model) and the field scale (2D radial model). Among the computed results are fluid loss vs. time, internal damage distribution, and productivity calculations for both the entire well and individual sections.The simulation model works, in part, independently of fluidloss experiments (e.g., the model does not use fluid-leakoff coefficients but instead computes the filter-cake buildup and its flow resistance from properties ascribed to the individual components in the mud). Some of these properties can be measured directly, such as particle-size distribution of solids, effect of polymers on fluid viscosity, and formation permeability and porosity. Other properties, which must be determined by tuning the results of the numerical model against fluid-loss experiments, are still assumed to be rather case independent, and, once determined, they can be used in simulations at altered conditions as well as with different mud formulations. A detailed description of the filter-cake model is given in this paper.We present simulations of several static and dynamic fluidloss experiments. The particle-transport model is used to simulate a dilute particle-injection experiment taken from the literature. Finally, we demonstrate the model's applicability at the field scale and present computational results from an actual well drilled in the North Sea. These results are analyzed, and it is concluded that the potential effects of the mechanistic modeling approach used are (a) increased understanding of damage mechanisms, (b) improved design of experiments used in the selection process, and (c) better predictions at the well scale. This allows for a more-efficient and more-realistic prescreening of drilling fluids than traditional core-plug testing. September 2010 SPE Journal Simulation ModelThis section describes the various modules that constitute the newly developed simulation model. The two main parts of the Maximize program are (1) the filter-cake model handling filter-cake buildup and controlling flow into the formation and (2) the reservoir-flow model handling flow inside the formation. The fluids introduced to the formation contain a number of dissolved and dispersed components, which, in turn, may change the original flow properties of the formation through various chemical and physical processes. The retention of solids and polymer is split into a pore-throat-trapping model and a pore-lining-adsorption model. Brine interaction with the rock surface and clays is described by a multicomponent cation-exchange mod...