Treatment of industrial mineral oil wastewater – effects of coagulant type and dosage

A coagulation dissolved air flotation (DAF) mechanism was employed in this study to evaluate the effects of four coagulants for industrial mineral oil wastewater (MOW) treatment. The investigation was done to improve the efficiency of a local South African oil refinery effluent treatment plant separating oil and water. The coagulants were aluminum sulfate (Alum), aluminum chloride (AC), ferric sulfate (FS) and ferric chloride (FC). Water quality parameters including chemical oxygen demand (COD), soap oil and grease (SOG), total suspended solids (TSS) and turbidity were used to evaluate the effectiveness of the coagulants at different dosages. Coagulant costs were compared at the optimum coagulant dosing rate and the most suitable coagulant selected for industrial MOW treatment. MOW is produced in high quantities from petrochemical industries and during crude oil refinery processes generates many oily pollutants that are harmful to the environment when untreated. In addition, an oily layer is created on water surfaces when MOW is discharged into them, increasing the COD and affecting aquatic life. This has led many researchers to try to find cost-effective physico-chemical processes to recover industrial mineral oil and water for reuse (Tsagarakis et al. 2003).

The production of mineral oil generates large volumes of wastewater, with low pH and high SOG. The high demand for water and oil associated with population growth generates numerous contaminants, hence treatment is necessary before discharge to protect public health and the environment. South Africa, as a fast-growing developing country, makes utilization of energy and water a great concern for social economic growth and sustainability. In addition, to improve fuel and water quality, due to the value of recovered mineral oil and the scarcity of water, environmental regulations are highly restrictive, with substantial sanctions and a requirement that effluent discharged to sewer must have SOG content below 50 mg/L (Welz et al. 2007; van Wilgen & Wannenburgh 2016). The recovery of valuable mineral oil is important for the environment, thus improving fresh water quality. According to Wang et al. (2012) mineral oil droplets in water exceeding 150 μm can be classified as free oil, dispersed oil with droplets between 20 and 150 μm, and emulsified oil with droplets less than 20 μm. The chemistry and location of the MOW affect the selection of both coagulants and treatment methods. However, MOW treatment is important for recovery and processing to yield lubricating oils and water, for industrial reuse and irrigation farming, and to protect ecosystems.

Coagulation is the most widely used chemical pre-treatment process for water and industrial wastewater, due to its economic and efficiency advantages, compared to alternatives, in reducing COD, SOG, TSS and turbidity (Altaher & Alghamdi 2011; Ngamlerdpokin et al. 2011). In coagulation, the coagulant is added to neutralize ionic charges to destabilize colloidal materials, causing small particles to agglomerate into larger, settlable flocs. However, coagulation and flocculation are interlinked, and require agitation to encourage the floc formed to agglomerate into larger masses in the solution. This involves rapid and slow mixing, where, in rapid mixing, the coagulants are dispersed uniformly, while slow mixing helps in inter-particle bridging of the floc size in the aqueous suspension. Many factors affect coagulation efficiency, including coagulant type and dosing rate, pH, mixing rate, and settling or flotation time (Ryan et al. 2008; Edzwald 2010; Rattanapan et al. 2011). According to Altaher & Alghamdi (2011) and Ngamlerdpokin et al. (2011), effective mixing occurs at between 100 and 300 rpm.

Studies have shown that DAF is the most effective application, employed with acidification and coagulation before discharge to wastewater agency facilities (Edzwald 2011; Rattanapan et al. 2011). However the coagulation DAF mechanism includes coagulation, flocculation and flotation, which depend on pH, and coagulant type and dosage (Daud et al. 2015). In acidification the pH of the MOW is changed to enhance the breaking of oil droplet emulsions. Lime, sulfuric acid, phosphoric acid, hydrochloric acid and sodium hydroxide are commonly used for pH adjustment. According to Daud et al. (2015), pH control operates in the coagulation DAF mechanism by generating oil droplet flocs and is highly efficient within the pH range 5 to 7. In addition, pH affects coagulation performance and pollutant removal thus oil in water under acidic condition can be easily be separated due to the apparent change in density. This also depends on water quality and the coagulant type – e.g., alum is effective over pH range 6 to 8 and AC in pH range 7 to 9 (Yang et al. 2010; Ngamlerdpokin et al. 2011). Work by Rattanapan et al. (2011) using coagulation in pH range 6 to 7 and AC at 1 g/L shows that 90% of SOG was removed from biodiesel wastewater. However at pH 5, an AC dosage of 2 g/L was used to obtain the same recovery rate. On the other hand, Yang et al. (2010) found that oil removal efficiency reached 96 and 59% at pH 4 and 9, respectively, using chitosan as an organic coagulant.

Many coagulants are used widely in conventional MOW wastewater treatment. They are classified variously as inorganic – e.g., alum, AC, FS and FC – synthetic organic polymers – e.g., zetag poly floc/Z553D, polyacrylamide derivatives, and polyethylene amine – and natural coagulants – e.g., chitosan, moringa oleifera, and bentonite clay (Ebeling et al. 2004; Rattanapan et al. 2011; Daud et al. 2015). Coagulant selection depends on economic consideration in wastewater treatment. This is an issue that has been well known in wastewater treatment since at least the mid-1970s. A review by Daud et al. (2015) suggested an overdose of coagulant could contribute to oil droplet restabilization, due to surface charge reversal on the particles and continuous chain reaction of the coagulants. To overcome these competing reactions, the effects of alkalinity, pH, trace elements and other compounds in the wastewater, the coagulant dosage required to remove SOG, COD, turbidity and TSS is usually established on the basis of jar tests, sometimes pilot-scale tests (Ebeling et al. 2004; Edzwald 2010). Experiments for this study were conducted using a DAF jar tester to evaluate the most effective and economic coagulant and its dosing rate, to improve the treatment efficiency and minimize the treated effluent's oil droplet content.

A MOW influent derived from mixed sources – refinery off-spec, ship slops, and refinery recycle streams – was collected from a local South African oilfield effluent plant. The influent is characterized in Table 1. The coagulants used were alum, AC, FS and FC, obtained from Sigma Aldrich.

To determine the suitable coagulant type and dosing rate, a 1 L stock solution was prepared for each coagulant. After acidifying the sample to pH 5, using 1.0 M H₃PO₄, the coagulant was added at doses of 0, 10, 20, 30, 40 and 50 mg/L to each 1 L sample, using a syringe in the sequence order. Each test was done at constant operating conditions – rapid mixing (250 rpm) for 2 minutes, flocculation (30 rpm) for 15 minutes, with a recycle ratio of 10% at retention time 3 seconds, air saturator pressure 350 kPa (50.7 psi), and flotation for 15 minutes. When flotation ended, 500 ml samples were collected for analysis. The same procedure was used for all coagulants.

Coagulant selection depends on the water chemistry and the quality demanded, coagulant dosage and economics (Ebeling et al. 2004). Both the Al³⁺ and Fe³⁺ based coagulants dissociate in solution, the trivalent ions enhancing polymeric chain reactions to adsorb oil droplets onto their surfaces to form flocs.

Figures 1 and 2 show that alum and AC are effective in removing SOG and COD, but that alum is superior due to a lower chemical consumption rate, with lower economic costs. Figure 2 also shows that increasing the FC dosing rate from 40 to 50 mg/L inhibited performance slightly in removing COD, from 84 to 83%. However, there was no significance effect on efficiency other than overdosing. Coagulant overdosing can lead to the restabilization of oil droplets, retarding the efficiency of the coagulant while increasing chemical costs (Daud et al. 2015).

Figures 3 and 4 also show that FC and AC were very efficient in removing turbidity and TSS. Because of the water quality requirements and environmental restrictions, FC is preferable as it gives better results at a lower dosing rate.

The optimum coagulant dosing rate was taken as 50 mg/L where the coagulants were more than 90% effective in reducing the oil content in water. Further, downstream, treatment would remove SOG to levels below the restriction limit. All of the coagulants have the potential to reduce the oil droplet content in water, but at a cost in terms of chemical consumption. Alum was most effective and economically viable, compared to the other coagulants.

Four coagulants were investigated in this study in order to select the most economic and effective for pre-treatment of industrial MOW. Coagulant selection based on performance depends on the key water quality parameters and the treatment process. In this context the study showed that:

• The best coagulant for pre-treatment of industrial MOW is alum, which is the most effective and economically viable in removing SOG from water.

• At 50 mg/L dose rate, FC is best for removing turbidity and TSS, while alum is best for the removal of COD and SOG.

• The coagulant dose increases the oil droplet floc size to enhance the efficiency of the process. Thus increasing the rate at which oil droplets combine with the floc, which then agglomerates and is precipitated.

• Coagulant overdosing could contribute to the restabilization of oil droplets and increase the cost of chemical usage.

The authors wish to thank FFS Refiners Research and Development Department, and Umgeni Water Process Evaluation Facility (PEF), South Africa, for their joint support in this project.

We also acknowledge the facilitators at a workshop on scientific writing, arranged by Young Water Professionals, South Africa, in cooperation with IWA, and supported by the Water Research Commission.