TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractThe Sognefjord reservoir (Jurassic) in Troll Field, Norwegian North Sea, is drilled horizontally in 9½" hole from 1,800 m to 3,100 m in length. The reservoir consists of loose sands, subdivided into clean and micaceous units and local hard calcareous cemented zones. These sections must be drilled in a tight horizontal plane, approximately one meter above the oilwater contact to maximize oil production. The oil bearing strata is typically between 12 m and 27 m thick. Most steering is done in the vertical plane to keep the well within the narrow target window (~1 m TVD or less) and in the optimum azimuth to place the well in the desired reservoir sand.Historically this section has been drilled with steerable motor assemblies and tungsten carbide insert roller cone bits. The use of polycrystalline diamond compact (PDC) bits was precluded due to problems associated with toolface control in the loose sands. With the advent of a rotary steerable tool (RSS) came the prospect of long runs with specialized PDC bits. After major area specific problems were encountered a joint task force was assembled, consisting of a drilling operator and an integrated service company, to attempt to improve drilling performance in the reservoir section.Three main areas of potential improvement were identified: 1. Improve reliability of the RSS (including MWD). 2. Increased ROP in hard calcite cemented sandstones. 3. Mapping and prediction of calcite cemented zones.The team concentrated on rock strength analysis, simulator testing of new PDC designs, BHA vibration modeling and analysis of depth based drilling, MWD & vibration data. Modified BHA's and Troll specific antiwhirl PDC bits, along with revised drilling practices, were applied in the field. This allowed step improvements in drilling performance to be achieved. Overall, meters/BHA increased from 264 m to ~640 m. The new bit designs also increased the overall average ROP from ~10 to 20 m/hr and improved instantaneous ROP in the calcareous sandstone sections. Savings in rig time on the Troll development are potentially immense as approximately 60 wells remain to be drilled. This paper will highlight the need for a systematic approach, including problem definition, data collection, solutions implementation and results benchmarking.