Final Report: Removal of Mercury and Other Heavy Metals of Industrial and Contaminated Site Waste Waters by Organic Chelation, Coprecipitation and High-Efficiency Particulate RemovalEPA Contract Number: 68D01062
Title: Removal of Mercury and Other Heavy Metals of Industrial and Contaminated Site Waste Waters by Organic Chelation, Coprecipitation and High-Efficiency Particulate Removal
Investigators: Hensman, Carl E.
Small Business: Frontier Geosciences Inc.
EPA Contact: Richards, April
Project Period: September 1, 2001 through September 1, 2003
Project Amount: $225,000
RFA: Small Business Innovation Research (SBIR) - Phase II (2001) Recipients Lists
Research Category: Watersheds , SBIR - Water and Wastewater , Small Business Innovation Research (SBIR)
The effluents of many industrial processes, as well as surface and groundwater from historically polluted sites, often contain unacceptably high levels of toxic trace metals. These must be removed before discharge to the environment. Although existing processes can treat most industrial waste streams, they often are costly, unpredictable, generate large volumes of secondary waste, and cannot meet the stricter standards recently proposed or considered by the U.S. Environmental Protection Agency (EPA). In July 1997, EPA released a summary report that evaluated all current commercially available Hg-removal technologies. None were able to meet even the most lenient discharge standards in place today. Frontier Geosciences, Inc.'s technology involves coprecipitation of the trace metal with a thiol-containing organic complexing agent (OTC). The reagent can be added to the aqueous sample in soluble form, and the complexed trace metals quickly separated from the resulting precipitate.
Limited acute aquatic toxicity testing of the OTC was conducted. Acute toxicity testing was carried out using Americamysis bahia (mysid shrimp) and Cyprinodon varigatus (sheepshead minnow). The acute testing demonstrated that the monomeric OTC would be rated as "slightly" toxic to marine aquatic organisms. However, it was determined that the polymeric OTC was environmentally nontoxic. Therefore, while the performance of the original monomeric OTC was monitored throughout the project, the polymeric version was used for field testing. To monitor the amount of OTC in the discharge of a treated wastewater stream, an assay needed to be developed. The resulting assay is simple, low-cost, and uses generic instrumentation found in any water testing laboratory.
It was discovered that the OTC preferably bound to lower valence state target metals, such as As(III) and Se(IV). Wastewater treatment failure often is a result of not understanding the individual wastewater chemistry, even in situations where novel and advanced treatment technologies are being used. The molecular and elemental speciation of the target metal can directly influence the efficacy of the treatment process being implemented. If the wastewater and treatment chemistry is known, a simple modification of the treatment process often yields the desired discharge concentrations. This is an important concept for the true understanding of how to correctly address a wastewater discharge problem, and one that often is ignored.
The competitive binding chemistry of the OTC is reasonably easy to deconvolve. From this, a binding hierarchy can be developed:
|Hg > Ag > Cu, Ni > Cd, Pb > Co, Se(VI) > Fe > As(III) > Zn > V > Tl > Cr|
|Most preferred||Least Preferred|
Also, it was found that ethylenediamine tetra-acetic acid (EDTA) could be used to mask the effects of ubiquitous metals such as Fe. The OTC prefers to bind to Fe over As(III), but EDTA preferentially binds the Fe. This allows the As(III) to be removed by the OTC. Interestingly, this reveals several questions when assessing the best strategy for treating wastewaters. For example, if As was a target metal for organo-thiol treatment strategies and the removal efficiency was not as expected, one must understand if the As was present as As(III), not As(V), and how much Fe was present in the wastewater. Identifying the correct situation would result either in pretreatment of the wastewater to convert As(V) to As(III) or the addition of EDTA. Also, if the target metal was Hg in the presence of a large concentration of Zn, the amount of binding reagent could be targeted in slight excess for Hg. This would compensate for process fluctuations, with the knowledge that there would be little ammonium pyrrolidine-dithiocarbamate (APDC) in the treated effluent because the Zn would scavenge the excess. In a groundwater situation where As(III) was the targeted metal, there would be a large concentration of Fe. However, an addition of the innocuous EDTA prior to organo-thiol treatment would assure that the As is precipitated from solution. However, the use of EDTA needs to consider the impact on the removal of the other metals in the groundwater.
In pure Hg-OTC, mercury is volatilized slightly at higher temperatures, regardless of purging in N2 or air. The Hg-OTC bond is stable in respect to heating in the pure compound, but does not seem to change to the HgS or m-HgS, more stable forms. Laboratory-treated waste containing mixed metals and organic compounds behaved differently than the pure Hg-OTC compound. Mercury losses by volatilization were significant upon heating. Other metals became more bioavailable upon heating, though there were no general trends of volatility or losses. The bioavailability of OTC-complexed metals, as assessed by the Toxicity Characteristic Leaching Procedure (TCLP), was lowest in wet OTC material with the notable exception of Hg. There was no general pattern of leachability among the metals when the treated laboratory waste was dried and heated. The methylation potential of the Hg-OTC complex, when introduced into sediments and soils, shows that Hg(OTC)2 is less bioavailable than HgCl2, but considerably more bioavailable than HgS. Plant uptake of the Hg-OTC complex was significant, but less than that of HgCl2. In summary, the material passes TCLP and is not bioavailable.
The effect of the organic contaminants often found in industrial wastewaters is dependent on the individual metal-OTC complexes. However, several generalizations can be stated: (1) for all metals tested, xylenes have the greatest impact on the metal-removal efficiency; (2) for all metals tested, the light petroleum hydrocarbon condensate has the least impact on the metal-removal efficiency; and (3) arsenic has a weak binding complex with the OTC and is the metal most greatly affected by the organic contaminants.
The particle formation of the OTC-metal complex was studied to ascertain the best separation strategy. This revealed several important considerations that pertain to the separation of the precipitation:
• Particle size distributions of systems containing "primary" particles and aggregates were bimodal or trimodal.
• The aggregation kinetics varied with metal and the aggregates also grow under quiescent aging.
• Reagent concentration, at a fixed metal/APDC ratio, had only a small effect on the aggregation kinetics. Increasing the OTC concentration sped up the aggregation process. Increasing the total dissolved solids speeds up the aggregation.
• Reducing the pH below 4.8 retarded aggregation; raising the pH only modestly increased the aggregation rate.
• The use of a coprecipitator for polishing allowed very low concentrations to be achieved.
• Polymeric OTCs had slower aggregation kinetics than the monomeric OTCs. Separation by flotation was as efficient as a 5 µ filter, with the use of a foam stabilizer.
A batch pilot plant was constructed and is still under study for the applicability to low-volume laboratory environments. At this point in time, the pilot plant has processed more than 500 gallons of high concentration toxic trace metal waste. The effluent is at least an order of magnitude below the local authorities’ discharge requirements.
Two client sites were characterized and tested for the applicability of OTC treatment. At the first site, a refinery, the total Hg concentrations in the wastewater are significantly reduced by the current wastewater treatment system. However, it is perilously close to the required discharge levels and any process event that disturbs the steady state of the current treatment system results in the discharge being out of specification. Between 60 and 99 percent of the Hg at the oil/water separator was present as a potentially volatile Hg species, essentially elemental Hg. The level could be reduced by a desanding hydro-cyclone. Unfortunately, arsenic cannot be reduced to the desired levels by a desanding hydro-cyclone. Concentrations of approximately 1 ppb Hg and 70 ppb arsenic can be achieved by the simple addition of polymeric OTC to the oil-loaded wastewater. This then utilizes the oil/water separator to remove the precipitate. If a coprecipitator is used at this point, Hg concentrations drop down to 0.006 ppb (6 ppt) and arsenic concentrations drop to 25 ppb, prior to the wetlands.
The second test site was the produced water line on a natural gas platform. A pilot desander was used to remove particulate Hg, bringing the initial Hg concentration down from approximately 300,000 ppb to about 200 ppb. Following the pilot-desander, a bench test determined that the OTC coprecipitation system reduced the produced water concentration of total Hg to 0.02–0.5 µg/L, depending on pH and use of a carrier. The OTC also was able to remove up to 99.3 percent of the arsenic, but only at reduced pH. For the next stage, a 1 gallon/min side stream of untreated produced water, on the natural gas platform, was connected to a pilot skid that had a mounted ceramic filter membrane. During the continuous-flow pilot testing using the polymeric OTC, Hg concentrations below detectable limits (0.2 ppt Hg) was achieved. Unfortunately, due to a well change, a significant amount of Fe appeared in the system and due to competitive binding, the arsenic concentration was not reduced. Given more time, EDTA as an additive would have been tested. Due to the success of the pilot study for removing the primary concern, Hg, the test site now uses a full-scale OTC binding system as the primary treatment method for 1,000,000 gallons per day of produced water.
The design of the continuous-flow pilot plant was meant to expand on the batch-flow system design described in the proposal, modified for optimal conditions and configurations determined in the other SBIR Phase II tasks. However, 12 months into the project, Frontier Geosciences, Inc.'s third-party sponsor indicated that they wanted to test the technology in a continuous-flow situation. This was exceedingly ahead of schedule. However, as described above, implementation was successful. Six other commercial sites have undergone onsite characterization, and the results are being interpreted as to the effectiveness of implementing an OTC treatment system. In some cases, it seems that a simple adjustment of their existing process, or cheaper existing technology, will solve their problems and the OTC technology will not be needed. However, the consulting income is still classed as incidental revenue applied to this project. This also is consistent with Frontier Geosciences Inc.'s belief that the client always should receive the most optimum solution and not have a single product forced on them when it is not needed. A dozen other companies have requested site characterizations, which are funded by the inquiring company, to determine whether this technology will help them. Interest in the technology has been garnered from other countries, such as Thailand, Australia, Finland, Spain, Czech Republic and Great Britain. It is without a doubt that Frontier Geosciences, Inc., will achieve its estimated market capture.
The project also has resulted in four papers currently submitted to peer-reviewed journals, as well as more than a dozen national and international presentations. Three other commercial ventures have been realized and implemented as a direct result of this project.