Perpustakaan Digital ITB

Heavy-oil wells of the Peregrino field form tight water-in-oil emulsions under the shear of the sand-control screens and the pump stages, amplifying the apparent viscosity up to eleven times the oil viscosity and choking ESP-lifted production. This study builds a validated nodal-analysis model of Well B in Prosper from published anchors (SPE-173972-MS) to quantify that penalty and to rank the operating responses against the equipment ratings. The emulsion enters as an apparent-viscosity table on the Pal and Rhodes form, whose single constant is fixed by two reported quantities, and partial demulsification is represented by a continuous excess-viscosity factor f, calibrated at f = 0.39 against the reported 45 percent field restoration. The model reproduces the reported design-basis operating point to within 1.2 to 7.0 percent, with one disclosed exception: the pump differential pressure is about 1.6 times the reported value, an offset that four vertical-lift correlations move by only 1.2 percent and that therefore lies in the reported data rather than in the modeling. The emulsion removes 60.5 percent of the deliverability at the 34.9 percent operating water cut and 76.7 percent at the inversion, and it leaves the design-basis completion unable to flow. The demulsifier gain is bell-shaped, peaking at 88.5 to 124 percent at the inversion and vanishing at both tails, so the chemical pays between roughly 30 and 50 percent water cut, a window that a screening economics widens from just above 10 percent. Of the operating conditions only the demulsifier is at once effective and feasible, since frequency is blocked by the 95 A cable above about 45 Hz and wellhead pressure is weak. Within the installed envelope no tested configuration is at once inside the ratings and productive under emulsion, so the response is either to resize the equipment or to remove the viscosity. The inversion points of the three reported wells collapse onto a log-linear trend that reproduces each well to within one percentage point when refitted without it, offered only as a screening relation within the 140 to 365 cP range from which it was derived.