Emerging Technologies

Phase II Summary Report – December 2011

Purdue University Calumet Water Institute – Argonne National Laboratory Research Group



BP has funded Purdue University Calumet (Purdue) and Argonne National Laboratory (Argonne) to identify deployable technologies that could help the BP Whiting Refinery meet wastewater discharge permit limits which will come into effect in 2012. The project is divided into two phases. Phase I, which started in the fall of 2007, focused on technologies for the treatment of total suspended solids and of ammonia in refinery wastewater, and was completed in June 2008. A copy of the Phase I final report titled “Emerging Technologies and Approaches to Minimize Discharges into Lake Michigan,” dated July 3, 2008, is available on the Purdue University Calumet Water Institute’s Web site. The project is currently in Phase II.

Summary of Phase II


Phase II focuses on the identification of technologies that are able to reduce the levels of vanadium and mercury in refinery effluent. It consists of multiple modules – with four modules either completed or in progress. The purposes of these modules are: to identify, list, pre-screen, select, and test technologies for the reduction of heavy metals, and to complete a comparative analysis of pollutant discharges to Lake Michigan.

Module 1, which commenced on February 1, 2009, comprised two major tasks:

  • to identify, list, and pre-screen technologies that could potentially reduce metals, in particular mercury, but also vanadium, selenium, and arsenic from refinery wastewater; and
  • to complete a comparative analysis of discharges into Lake Michigan that was started in Phase I of the project.

The task of screening technologies, Module 1, was completed on May 31, 2009, and the results are summarized below. A report on the comparative analysis task was submitted separately, and will be published at the Purdue Calumet Water Institute website by the first-quarter of 2011.

Purdue and Argonne researchers independently reviewed technical information on metals treatment technologies provided by technology vendors, and from patents, open literature, academia and other sources. For each technology found, this study reviewed publicly available data on testing, technology stage of development, and specifications. Removal efficiency was assessed for mercury and vanadium. The probable costs, spatial requirements, and residual management of each metals removal system were considered to the extent information was available.

Limited information was found regarding the treatment of refinery wastewater. Most of the tests whose results were reported in literature and by vendors were not conducted with refinery wastewater; hence removal efficiency may be different than that reported when refinery wastewater is tested. Furthermore, the concentrations of heavy metals in the wastewater reported in literature and by vendors in most cases were higher (often by several orders of magnitude, particularly for mercury) than those generally found at Whiting. Another issue in the reported data for mercury removal technologies is the relatively high detection limits – much of the data was apparently generated before the new Method 1631 was finalized, or was collected with other mercury analytical techniques.

The researchers have prepared a report that summarizes the information found as a result of this work. This report has identified several processes for metals removal from wastewater streams such as precipitation, adsorption, ion exchange, biological/hybrid treatment, membrane filtration, chemical reaction, and electrical/magnetic processes. A few technologies were found that show promise in the achievement of the required low-level concentrations in industrial wastewater.

Module 2 included the sampling and analysis of a number of waste streams at the Whiting Refinery. This characterization study ended in June 2009, and served as the experimental design basis for technologies tested in Module 3.

Module 3 started in September 2009, and was aimed at:

  • providing comparable, transparent and uniform bench-scale test data for a range of wastewater treatment technologies identified from Module I to reduce  mercury, selenium, arsenic, and vanadium.
  • recommending technology(ies), if any are found suitable, for pilot-scale testing

The researchers have conducted over 30 bench-scale treatability studies in seven technology categories of co-precipitation, adsorption, ion exchange, membrane filtration, chemical reaction, sequestration of mercury, and electrical/magnetic processes. For mercury, the clean-hand sampling procedure was employed for wastewater sample collection, and standard protocols were followed for the other metals. All of the technologies for treatment of mercury were tested independently at Purdue and Argonne under “clean room class 100” laboratory conditions to minimize background contamination. Also, one technology was tested under standard laboratory conditions at Purdue. Other investigations focused on understanding the chemical form and species of metals in the wastewater to better understand technology performance.

For each study, the initial and final metal concentrations, and method and equipment blanks were analyzed by an independent laboratory using appropriate USEPA method. The removal efficiencies were evaluated and used for comparative assessment of the technologies. Technologies that could not meet the expected discharge permit limits were excluded from further consideration for pilot-scale testing.

The subset of tested technologies that were able to meet the discharge limits in the bench scale tests were further evaluated for technology readiness for full-scale applications as well as other performance criteria such as power consumption, cost, and waste generation to the extent these parameters can be measured or derived. Because of scalability issues from bench-scale to pilot/full scale, the energy, cost, and secondary waste generation data at the bench scale may not be truly representative. The lab study found that promising technologies and potential treatment alternatives potentially could be used to reduce mercury and vanadium discharged into the Great Lakes by industry. The researchers prioritized these technologies for further testing:

  • Ultrafiltration because it is effective for the removal of particulate mercury in the lab.
  • Adsorption because it is effective for the removal of dissolved mercury.
  • Reactive filtration because it is an emerging technology that could be useful for treating both particulate and dissolved mercury.

The Purdue and Argonne researchers have prepared a comprehensive report of their findings from the pre-screening of technologies (module I) and the treatability studies (module III). These reports were reviewed on October 2010 at Purdue University, Calumet campus by independent external experts from academia, government, and industry. This peer review process provided an unbiased and objective scientific assessment of the results and conclusions of this study. Phase 2 Modules 1-3 reports are available on the Purdue University Calumet Institute’s web site.

Module 4 started in May 2011, and is aimed at:

  1. Confirming bench-scale performance on achieving concentrations less than 1.3 ppt mercury and determining removal of other analytes specified in the work plan.
  2. Demonstrating and evaluating performance under continuous feed conditions and variable wastewater composition

Pilot study of the ultrafiltration membrane and reactive filtration systems at BP Whiting Refinery has been performed to verify technologies performance under varying conditions on a larger scale. Analysis of the data acquired during the pilot study is ongoing.