In the development of Elf Exploration Angola’s Girassol field located approximately 200 km west of the coast of Angola in 1,500 m of water depth, 390,000 barrel per day of seawater will be needed for a waterflood operation. Formation water analysis indicate both barium and strontium levels of approximately 230 mg/l. By the injection of normal seawater containing 2,800 mg/l sulphate, scale tendency calculations indicate the possible formation of more than 320g barium sulphate per m3 of coproduced water. Conventional scale inhibitor squeeze treatments were not considered to be a reliable and cost effective scale control technique due to: 1) Girassol’s "daisy chain" concept with network of wells producing in the same gathering line and subsequent difficulties in monitoring residual inhibitor levels, 2) difficulties in the placement of inhibitors in the reservoir which has high permeability and long perforation intervals, and 3) the deep and remote subsea wells. The high risks using scale inhibition for such a large scale waterflood project in a deepwater environment were considered unacceptable. After the extensive screening of all possible technical alternatives, Elf chose to remove the sulphate from the seawater using a selective nanofiltration membrane process previously used in the North Sea1,2 . However, unlike the North Sea operations where scale inhibitors were originally used in combination with low sulphate seawater, Elf chose a specifically designed sulphate removal process with stringent high performance requirements that allows for the initial injection of seawater with approximately 20 mg/l sulphate for the leading edge of the waterflood. This will be followed by low sulphate seawater with approximately 40 mg/l sulphate. Based on 1) the initial low sulphate level in the injection water, 2) the projected scaling kinetics, and 3) the partial precipitation of barium and sulphate in the reservoir before the seawater approaches the production bore, the calculated scale potential is essentially eliminated.