Title : Improving oil removal from produced water: A comparative study of granular media in depth filtration
Abstract:
Produced water is a major by-product of oil production that contains residual oil and suspended solids, requiring effective treatment to meet stringent standards whether for reinjection or discharge to the environment. Depth filtration – a typical polishing step that removes emulsified and remaining free oil - was investigated using various granular media for oil removal from synthetic produced water (sPW). The sPW was formulated with two types of crude oil (A or B, at 50 ± 5 mg/L total oil and grease (TOG)) dispersed in brine (40 g/L NaCl and sodium dodecyl sulphate (SDS) at concentrations of 2 mg/L for Oil A and 4 mg/L for Oil B). The droplet size distribution of the emulsified oil was characterized (Dv10 = 1.6 µm, Dv50 = 9.8 µm, Dv90 = 28 µm) to ensure compatibility with industrial conditions. Laboratory tests were conducted at a filtration rate of 30 m/h using a 50 cm bed of one pre-washed media, including walnut shell (NS), macaw palm endocarp grit (EG), zeolite medium (ZM), and diatomaceous earth (DE). The performance of each granular medium was evaluated by monitoring the outlet TOG concentration via cyclohexane extraction followed by visible light spectrophotometry at 400 nm. Filtration runs were performed until significant performance loss occurred, ranging from 12 to 48 hours. The removal performance was segmented into stable – further divided into best, and good periods – and unstable periods, providing insight into the operational stability and efficiency of each medium. Additionally, backwash outlet TOG was monitored after the end of each filtration run. Results indicated that for Oil A, walnut shell maintained stable performance over a longer time, while diatomaceous earth yielded the lowest effluent TOG during the best performance period, although its long-term stability was comparable to that of the zeolite medium. The endocarp grit demonstrated unfeasibility due to an insufficient stability duration. In contrast, with Oil B, NS, DE and ZM exhibited similar outlet TOG, whereas the duration of the best performance period was consistently longer than with Oil A. Variations in backwash efficiency further underscored the influence of oil type on system performance. The initial backwash kinetics (outlet TOG x time) was modelled with exponential regression (R² ≥ 0.94), and these results were correlated with the total oil mass retained during filtration, being sensitive to the oil type. In conclusion, the study demonstrates that selecting granular media based on the quality of produced water is crucial for optimizing produced water depth filtration processes in oil production. The findings provide valuable insights for the design and operation of treatment systems, suggesting that tailored media selection can enhance oil removal efficiency, extend filtration cycles, and reduce operational costs, improving the sustainability and economic feasibility of produced water treatment in the oil and gas industry.
