Effect of Water Chemistry on Polymer Retention in Carbonate Reservoirs under Harsh Conditions

Abstract

Polymer flooding is one of the most effective chemical EOR method and has been successfully applied in many sandstone reservoirs worldwide. In this process, a water-based solution containing long-chain polymer molecules is injected into the reservoir to reduce the mobility ratio between the injected fluid (water) and the displaced fluid (oil); thus, improving the sweep efficiency and increasing the amount of oil recovered. In recent years, polymer flooding has emerged as a promising enhanced oil recovery technique in carbonate reservoirs of the Middle East due to the development of novel polymers that can withstand the harsh conditions in these reservoirs. One of the significant challenges of polymer flooding is polymer retention affecting the economics and the performance of a polymer-flooding project.

The main objective of this research is to investigate the effect of water chemistry (salinity and ionic compositions) to reduce polymer retention in carbonates at representative reservoir conditions. SAV10, an ATBS based polymer that showed significant stability at high temperature and salinity, was considered. The water chemistry of the make-up brine was altered by reducing the salinity of both typical formation water and seawater in carbonates under harsh conditions. Furthermore, varying the ionic composition of Ca2+, Mg2+, and SO4 2- was also investigated to reduce polymer adsorption. A significant aspect of the study was the assessment of polymer adsorption when oil was present, and the influence of temperature on polymer retention. Furthermore, polymer adsorption was quantified using both static adsorption tests on crushed Indiana limestone outcrop and dynamic retention test on the same Indiana limestone outcrop core plugs. Finally, a correlation was developed between static adsorption and dynamic polymer retention.

Twelve proposed brine recipes including both water dilution and ion manipulation effects were selected. The rheology of the polymer was closely studied, emphasizing shear ramp-up tests, varied polymer concentration, salinity, ionic compositions, and temperature to study their effects on polymer solution viscosities. The hydrodynamic size and polydispersity index of the polymers were determined at varying conditions. Various analytical methods including UV-Vis, TOC-TN, and viscosity, employed for determining polymer concentrations were compared to identify the most accurate and consistent technique. Static and dynamic retention studies were conducted to study the effect of salinity, ionic compositions, and temperature on polymer retention. Moreover, additional static adsorption experiments were performed at the same experimental conditions including surface access, liquid-to-solid ratio, retention time, make-up water salinity and temperature as that of dynamic retention tests. Based on these results, a correlation was developed between static adsorption and dynamic retention with the application of Machine Learning algorithms and symbolic regression.

To investigate the effect of salinity on polymer retention, dilutions of formation water salinity (167,114 ppm) and seawater salinity (42,507 ppm) were first studied, which are representative of a Middle East reservoir. Initially, static adsorption studies were conducted as the preliminary screening study. This was followed by dynamic retention studies in the absence of oil as well as in the presence of oil on oil-wet cores to obtain representative polymer retention values. All the experiments were performed at ambient temperature conditions. Both static and dynamic polymer retention studies showed reduced retention levels below a salinity of 10,000 ppm in the absence and presence of oil. Moreover, in the presence of oil, the polymer retention was less compared to its absence. The dynamic polymer retention was 14 µg/g-rock for seawater dilution of 425 ppm salinity compared to 26 µg/g-rock for high salinity formation water (167,114 ppm) in oil wet cores.

Furthermore, the effect of ionic compositions on polymer retention was studied using brines derived from seawater of salinity 8,502 ppm with varying concentration of Ca2+, Mg2+ , and SO4 2- . The investigation included static adsorption studies as well as dynamic retention experiments in the absence and presence of oil at ambient conditions. A reduction in static and dynamic polymer retention was observed with reduced hardness levels. The dynamic polymer retention for complete softened brine (0 ppm of Ca2+ + Mg 2+) in the presence of oil was 20 µg/g-rock, whereas the dynamic retention value for the brine composition with hardness ions (425 ppm of Ca2+ + Mg2+) was 26 µg/g-rock. Analyzing the low polymer retention values obtained based on water dilution and ion-modification, diluted brine compositions showed lower retention (14 µg/g-rock) compared to softened brine composition (20 µg/g-rock) in the presence of oil. When considering large-scale applications, it is simpler to formulate a diluted brine recipe than to vary ionic compositions at the field scale. Hence, diluted brine composition was recommended as preferred solution.

In high-temperature conditions, the study investigated the static and dynamic retention of polymer in low salinity brine (425 ppm) compared to high salinity (167,114 ppm). Static adsorption showed decreasing polymer adsorption with increasing temperatures in both salinity levels. However, dynamic polymer retention remained stable across various temperatures (25, 40, 60, 80, and 90°C) for both salinities, attributed to the stability of polymer at high temperature. Despite this stability, the polymer's hydrodynamic size reduction at high temperatures leads to increased retention. On the other hand, high temperatures weaken the intermolecular forces between polymer and rock, reducing retention, which is more apparent in static adsorption due to greater surface area. In dynamic coreflooding, where surface area is less, these effects balance out, resulting in minimal temperature impact on retention. At 90°C, mimicking reservoir conditions, polymer retention in low salinity (425 ppm) was 12 µg/g-rock.

For effective polymer flooding design, precise polymer retention measurement is crucial. The final objective aimed to correlate static and dynamic retention values. Using crushed rock powder, static adsorption experiments were conducted, ensuring similar surface access to that of the core plugs used for dynamic retention studies. Predicting dynamic retention involved machine learning and symbolic regression, initially focusing on static adsorption values. All tree-based models exhibited superior performance in predicting dynamic retention demonstrating R2 values of 0.95 – 0.98 and 0.91 – 0.95 for training and testing, respectively, with respective mean absolute error (MAE) and root mean square (RMS) error values in the range of 6.13 – 8.17 and 8.10 – 9.83. Further analysis included variables like TDS, pore throat size, residual oil saturation, and temperature. This holistic approach, encompassing more parameters, improved accuracy, as evidenced by high R2 values (0.98 – 1.00 for testing, 0.97 - 0.98 for training) and lower MAE and RMS ranges of 2.97 – 4.26 and 3.71 – 5.31, respectively. Symbolic regression, applied for the two distinct approaches, yielded explicit equations with R 2 values of 0.95 and 0.98, though the latter had a higher RMS (19.82 vs. 6.84). Nevertheless, the second approach using different influential variables captured a broader range of influencing factors, rendering it more valuable for practical applications.

This research is among the very few works that investigate the effect of injection water chemistry, including both water dilution and ion-modification, on polymer adsorption in carbonates under harsh conditions of high temperature and high salinity. Furthermore, the developed correlation could determine representative polymer retention levels at reservoir conditions based on static tests, which save the time and efforts spent in dynamic retention tests.

Date of Award	15 Dec 2023
Original language	American English
Supervisor	Emad Al Shalabi (Supervisor)