An experimental study of polymer flooding is presented here, focusing on the influence of initial core wettability and flood maturity (volume of water injected before polymer injection) on final oil recovery. Experiments were performed using homogeneous Bentheimer Sandstone samples of similar properties. The cores were oilflooded using mineral oil for water-wet conditions and crude oil (after an aging period) for intermediate-wet conditions; the viscosity ratio between oil and polymer was kept constant in all experiments. Polymer , which is a partially hydrolyzed polyacrylamide (HPAM), was used at a concentration of 2,500 ppm in a moderate-salinity brine. The polymer solution was injected in the core at different waterflood-maturity times [breakthrough (BT) and 0, 1, 1.75, 2.5, 4, and 6.5 pore volumes (PV)]. Coreflood results show that the maturity of polymer injection plays an important role in final oil recovery, regardless of wettability. The waterflood-maturity time 0 PV (polymer injection without initial waterflooding) leads to the best sweep efficiency, whereas final oil production decreases when the polymer-flood maturity is high (late polymer injection after waterflooding). A difference of 15% in recovery is observed between early polymer flooding (0 PV) and late maturity (6.5 PV). Concerning the effect of wettability, the recovery factor obtained with water-wet cores is always lower (from 10 to 20%, depending on maturity) than the values obtained with intermediate-wet cores, raising the importance of correctly restoring core wettability to obtain representative values of polymer incremental recovery. The influence of wettability can be explained by the oil-phase distribution at the pore scale. Considering that the waterflooding period leads to different values of the oil saturation at which polymer flooding starts, we measured the core dispersivity using a tracer method at different states. The two-phase dispersivity decreases when water saturation increases, which is favorable for polymer sweep. This study shows that in addition to wettability, the maturity of polymer flooding plays a dominant role in oil-displacement efficiency. Final recovery is correlated to the dispersion value at which polymer flooding starts. The highest oil recovery is obtained when the polymer is injected early. Introduction Enhanced-oil-recovery (EOR) technology is receiving increasing attention by many companies as a solution to satisfy the global oil demand. One of the most implemented chemical-EOR technologies is the injection of polymer, which has been used successfully for decades. Among all the polymers proposed for EOR applications, HPAM is the most widely used polymer for polymer floods. HPAM is a viscoelastic polymer; its viscosity varies with shear rate and its rheological behavior depends on molecular weight, hydrolysis degree, concentration, ionic strength of brine, temperature, and pH. The main objective of polymer flooding is to decrease the water mobility by increasing its viscosity as well as decreasing water relative permeability to improve macroscopic oil-sweep efficiency (Lake 1989; Wassmuth et al. 2007; Koh et al. 2016). This injection can be performed as a secondary polymer flood or after an initial waterflooding period (tertiary polymer flood), when the water cut reaches high values. Secondary and tertiary polymer floods, however, behave differently. Cottin et al. (2014) show a lower remaining oil saturation resulting from secondary polymer floods compared with tertiary polymer injection, as a result of a better sweep efficiency at the pore scale. The efficiency of a polymer flood depends on several factors that should be considered before starting the injection. Skauge et al. (2014) and Shashvat and Mohanty (2015) showed the importance of the viscosity ratio between an oil and brine/polymer solution for the development of water channels (viscous fingering) and its effect on water BT (WBT) and oil recovery. Their results show that the number of fingers is reduced with an increase in the polymer viscosity, and consequently the sweep area is also increased. A very low concentration of polymer (low viscosity) might in some cases bring an important incremental oil recovery (Levitt et al. 2013). Viscoelastic effects of polymer solutions have also been shown to decrease oil saturation over purely viscous fluids (glycerine). Under the same conditions, Wang et al. (2000), Vermolen et al. (2014), and Qi et al. (2016) conducted a series of corefloods and observed a clear reduction of oil saturation as a result of the viscoelastic effect of the polymer. However, for high-viscosity crude oil, this viscoelastic effect did not improve recovery over polymer floods with low viscoelasticity. The capillary number is another important parameter; Shashvat and Mohanty (2015) showed the effects of flow rate on fingering and oil recovery. At the same viscosity ratio (l o = l w ¼ 1; 000), the fingering effects are more important at high flow rates, causing a reduction of oil recovery. Core permeability (Asghari and Nakutnyy 2008) also affects the oil recovery obtained under similar operational conditions (injection rate and viscosity ratio), and is higher in the case of high-permeability cores than in low-permeability cores. Finally, wettability is another important parameter that plays a key role. Oil-displacement efficiency is controlled by the phase repartition at the pore scale and by the flow instabilities during waterflooding at the core scale. It is known that the remaining oil saturation obtained after waterflooding depends on wettability (Jadhunandan and Morrow 1995); the lowest value is obtained with intermediate-wettability cores, but there is no similar evidence for polymer flooding. Shiran and Skauge (2015) showed the influence of wettability on resistance and the residual factor (mobility ratio). A lower extra oil-recovery factor was obtained as a result of a higher resistance and residual factor in water-wet systems than in intermediate-wet systems.< Réduire