Benzene Removal in a Petroleum Refinery Moving Bed Biofilm Reactor System

Executive Summary A study was performed to evaluate benzene removal and biomass concentrations in two parallel moving bed biofilm reactors (MBBRs) at the Valero Ardmore Refinery, Ardmore, OK (the Refinery). Three rounds of daily testing were completed between July and November 2022. Study results indicated greater than or equal to 99.8 percent benzene biodegradation across the MBBRs. Effluent benzene concentrations were generally below the analytical limit of detection. The average MBBR biomass concentration was 478 milligrams per liter (mg/L) as total suspended solids (TSS), with an approximate 50:50 split between biomass in the bulk liquid and that attached to the MBBR media. Study results led to the State of Oklahoma making a regulatory determination that the MBBRs as operated by Valero meet the definition of an 'Enhanced Biodegradation Unit' under the federal Benzene Waste Operations National Emissions Standard for Hazardous Air Pollutants (BWON). Thus, going forward, air emissions controls are not required for the MBBR off gas. Introduction The Refinery operates a wastewater treatment plant (WWTP) to manage process wastewater, with final effluent discharged to the Washita River. The WWTP includes two parallel bioreactors, former aeration tanks converted to MBBR technology circa 2006 with the addition of 328 cubic meters (m3) of AnoxKaldnes K1 media to each unit. In early 2022, an additional 160 m3 of World Water Works WWW-05 media was added to Reactor 2. Both media types have 500 m2/m3 of specific surface area. Treatment of the Refinery's wastewater is subject to 40 CFR Part 61 Subpart FF, also known as BWON. This rule allows for an exemption from air emission controls for bioreactors meeting the definition of enhanced biodegradation. As MBBR technology does not recycle biomass, it was originally believed that the Refinery's bioreactors did not qualify for the BWON exemption. Accordingly, the MBBRs were designed with covers and a thermal oxidizer to control air emissions from the combined vent gas. This study was performed to evaluate the fate of benzene in the MBBRs, with the goal of demonstrating that vent gas control could be eliminated while satisfying BWON requirements. Three rounds of testing were performed around each reactor to collect the following data: Influent benzene, Effluent benzene, Benzene in vent gas, Amount of MBBR media in each reactor, Biofilm mass on the MBBR media, TSS. The data obtained allowed for calculation of percent benzene biodegradation and total biomass quantity in each MBBR. Methodology Table 1 summarizes sample locations and parameters analyzed during each sampling round. Vent gas samples were collected using evacuated Summa cannisters equipped with flow regulators that provided for sample collection over a 1-hour period. The sampler tubing was connected to an existing tap on the common vent gas header from the two bioreactors, upstream of off-gas treatment. The wastewater samples were collected from existing sample taps. Media samples were taken from the aerated bioreactors using a Kemmerer sampling device. To quantify the attached biofilm growth on the MBBR media, the World Water Works procedure was used. The percent media fill in each reactor was estimated based on counting the number of individual media pieces collected per Kemmerer sampler grab. Results Table 2 contains wastewater benzene results from the three rounds of testing. Influent benzene concentrations were 6,950 to 10,900 micrograms per liter. Effluent benzene concentrations were generally non-detectable (detection limit of 0.33 µg/L or 0.66 µg/L), apart from one measurement of 1.33 µg/L on the Round 3, Reactor 2 effluent sample. Table 3 shows vent gas benzene results. Although results varied somewhat between the two MBBRs and among the different sampling rounds, values were low in all cases, with a maximum benzene content of 960 µg/m3. Table 4 summarizes results of the TSS and volatile suspended solids (VSS) analyses performed on wastewater samples. Data presented are the averages of triplicate determinations. Table 5 provides results of biomass analyses on the MBBR media. The biomass on K1 media was reasonably consistent between the two reactors, averaging 3.42 grams per square meter (g/m2) for Reactor 1 and 2.76 g/m2 for Reactor 2. The amount of biomass was also reasonably consistent between the two types of media in Reactor 2 when normalized to specific surface area, 2.76 g/m2 for K1 and 2.27 g/m2 for WWW-05. It was estimated that Reactor 1 had a 15.1 percent media fill and Reactor 2 had 18.9 percent fill. To calculate the benzene mass balance around the bioreactors, wastewater and vent gas flow rates were available through existing on-line flow meters. Table 6 summarizes benzene removal by biodegradation based on mass balances. Benzene biodegradation was greater than or equal to 99.8 percent for each of the testing rounds. Table 7 summarizes combined TSS results to quantify the total biomass in each reactor during each round of testing. Total biomass concentration averaged 478 mg/L TSS. On average, 54 percent of the biomass is biofilm attached to the media, and the balance is suspended growth in the bulk liquid.