Another Tool in the Remediation Arsenal?

The use of in-situ chemical oxidation (ICO) for the treatment of organic environmental contaminants is an emerging technology. There are several oxidants currently being studied and used to treat chemicals in process streams and to remediate sites affected by various chemical contaminants. Among the various oxidants being used are hydrogen peroxide (H2O2), ozone (O3), sodium permanganate (NaMnO4), potassium permanganate (KMnO4), oxygen and Fenton’s reaction chemistries. The ability of these chemicals to treat contaminants in a matter of minutes, days and weeks as opposed to months or years for other remediation technologies has generated interest in studying the effectiveness and impacts of their use.

The relatively widespread use of potassium permanganate in the chemical industries and for wastewater treatment has been ongoing for several decades. The KMnO4 reacts with and oxidizes the organic carbon in the treated media. Among the resultant by-products are potassium ions, manganese dioxide (MnO2) and carbon dioxide (CO2), as well as other less complex organic compounds that are less hazardous and more easily degraded through natural processes.

The addition of MTBE to gasoline and the resulting groundwater impacts from leaking underground fuel storage tanks (LUST) has been a major issue on regulatory agendas and in the news over the last several years. The need to identify, control and treat these release sites in a timely, cost effective and environmentally sound manner is a matter of concern throughout the United States. Once a release is identified and the source of the contamination (e.g. leaking tank, broken fuel transfer line, etc.) is corrected, a remediation technology that could rapidly oxidize the gasoline components would be a very useful tool. Second only to prevention of fuel releases, rapid source treatment is the most effective way to prevent adverse, long-term impacts to groundwater resources.

Components of fuel releases include, but are not limited to aromatic hydrocarbons (e.g. benzene, ethylbenzene, etc.) and fuel oxygenates (e.g. MTBE, TBA, etc.). It is well known that benzene is naturally attenuated readily once the source of additional contamination is stopped. Natural attenuation in advancing groundwater plumes is well documented for aromatic compounds. The ability of natural attenuation to mitigate MTBE has been demonstrated at several sites. What should not be lost on these debates is the effective and rapid reduction of the concentration of chemicals in the source area is the most critical factor in protecting groundwater. The nature of these compounds and the relatively localized initial soil impacts at most LUST sites are good candidates for treatment by ICO.

MTBE Bench Scale Groundwater Treatability Study

Several studies at various sites have shown that KMnO4 is effective at dramatically and quickly reducing contaminant levels of recalcitrant organic compounds including tetrachloroethene, trichloroethene, as well as other chlorinated and hydrocarbon compounds. The reaction chemistry is well documented for many types of contaminants across a wide range of pH and other water quality parameters.

There has not been much publicized on the effectiveness of KMnO4 to treat MTBE as part of a detailed bench or field scale treatability study. To address this, bench scale treatability studies were performed on groundwater extracted from an active chemical manufacturing plant site. The test was performed to determine the effectiveness of various doses of KMnO4 on the concentration of MTBE and other contaminants in the groundwater. Groundwater was placed in 40 ml. VOA vials and 3% KMnO4 solution was added to attain 3 different effective concentrations. Two concurrent replicates were performed to confirm results. The samples were analyzed after 3, 10, 17 and 24-day treatment periods to determine treatment time required using varying oxidant dosages. The samples were analyzed at various dilutions by a modified method SW846-8260B.

The initial concentration of MTBE in the groundwater was 11500 ug/L. Other compounds in the initial groundwater sample included TBA at 50000 ug/L and benzene, ethylbenzene and a few other compounds at varying concentrations under 500 ug/L.

The results of this test clearly indicate that KMnO4 is effective at degrading MTBE and that differing dosages of KMnO4 directly impacts the rate and extent of the oxidant reaction. The last sample analyzed on the 24th day of the test did not have any detected compounds other than MTBE and TBA. No evaluation was done on the concentrations of non-target list compounds, or tentatively identified compounds (TICs) detected in the sample. The target list for these analyses included the standard EPA-CLP compound list plus TBA and MTBE. It is probable that some of the TICs detected in the sample were breakdown products of MTBE or other compounds in the initial sample.

The analytical values for TBA were extremely variable for this test. It is reasonable to assume that some of MTBE was oxidized into TBA as part of the reaction pathway to eventual complete oxidation. Because a purge-and-trap method (modified method SW846-8260B) was used to analyze these samples, issues such as purge efficiency and carry-over accounted for some of the analytical variability. In retrospect, confirmation of TBA concentrations by a direct aqueous injection method would have provided more definitive analytical data. Based on the analytical results of this bench scale test, no conclusions could be drawn on the ability of KMnO4 to oxidize TBA.

Technology Limitations

In the case of ICO, the limitations are critical not just to the success of the effort, but also to preventing unforeseen long-term problems. Anytime a substance is added to a site where the problem is a more hazardous substance, there is an apprehension factor that must be overcome. The questions that may arise aren’t hard to imagine:

• Is the substance we are adding hazardous in and of itself?
    
• Will the additive degrade or oxidize into other hazardous forms?
    
• Will the additive dissolve or otherwise liberate other substances that are not currently an issue at the site?
   
• What are the long-term impacts of the additive on groundwater resources?
    
• Will the oxidant by-products (colloidal solids )plug the pores in the aquifer?
    
• Will the act of adding this substance create long-term liability for companies or contractors?

Because ICO is an additive technology, a variety of issues must be considered before implementation. Among the issues are hydro-chemistry, hyro-geology, nearby receptors, and other inorganic substances, such as heavy metals, that may be present. In this regard, all remediation technologies involve the evaluation of many of these same factors.

The reluctance to use a new technology such as ICO is understandable when reading through the cautionary statements above. Technological limitations are not unique to ICO or any other remediation technology. The limiting factors may involve effectiveness, timeliness, by-product generation, cost, complexity, proximity to immovable infrastructure or receptors, availability and many other issues. In the case of LUST sites and many other types of remediation projects, the concept of a using a series of techniques and/or a variety of technologies is common. While one part of the process may involve digging up the leaking tank, piping or soils, follow-up treatment may be required in the form of thermal or chemical oxidation of heavily impacted, but less easily accessed soils. Further efforts may also be required to assist or enhance natural attenuation in the groundwater contaminant plume through the use of air sparging or injection of electron acceptors such as oxygen, hydrogen or nutrients.

As with any remediation technology, the positives must be weighed against the negatives before a decision is reached as to how to address a given problem. Among the benefits of using ICO technology are:

• Relative ease of system design. There is usually no need for a large-scale groundwater treatment or fugitive vapor abatement system. ICO systems using oxidants that cause exothermic reactions, such as Fenton’s reaction systems, may require vapor control. Detailed site definition in terms of contaminant concentration and location are common to any technology selected. Calculation of the mass of oxidant needed to oxidize contaminant mass and delivery method is a primary design consideration.
     
• Time frame is much shorter than most technologies. Depending on the permeability of soils and mass of contaminant at a site, the reaction time is usually measured in terms of several weeks. Soil vapor extraction, in-situ bioremediation and oxygen sparging project time frames are measured in months or years.
     
• ICO is a destruction technology. Other remediation technologies may focus on transferring the waste from one phase (aqueous) to another phase (vapor) before treatment. Digging up waste soil is certainly appropriate in some instances, but the waste (and liability) is just getting moved to a different location.
    
• ICO can be implemented around immovable infrastructure. There is usually no need for extensive piping and other long-term remediation system components (pumps, compressors, frac tanks, etc.) to remain on-site.
     
• Progress of ICO can be monitored visually. When KMnO4 solution reacts with organic carbon, the solution turns from a brilliant purple color to brown, amber and eventually to clear when all of the oxidant has reacted.
    
• Overall costs of using ICO appear to be very competitive with alternative technologies. The different oxidants used vary drastically in cost (NaMnO4 is at least 5 times more costly than KMnO4). Overall savings are a result of lower design, installation, long-term operation and infrastructure costs. The use of this technology is relatively new and available project cost data is very limited. Although little data is available on overall costs of ICO projects, shorter project time frames usually translate into overall cost reductions.

Delivery Methods

There are several delivery methods being tested and implemented. The two most common methods are injection of oxidant through a well or gallery of wells and by using a lance. A lance is essentially a one-time procedure for delivering a substantial quantity of oxidant over a short period of time to discreet vertical and lateral locations. Injection wells are used for delivering smaller quantities or concentrations of oxidant over a longer time frame usually to a larger target zone. Groundwater extraction wells may be used to help distribute the oxidant throughout the target zone. The following graphic illustrates the concept of lance injection delivery.

With a well-defined site, the oxidant delivery program can be designed such that oxidant/contaminant contact can be optimized. An oxidant delivery lance is driven to specific depths over a given grid pattern and a volume of oxidant is injected at vertical intervals dependent on location of contaminant. The KMnO4 solution is slightly denser than water and flows downward and down gradient of the injection point (see graphics above). This density driven factor has been shown to significantly oxidize DNAPL that settles on aquitards and low permeability clay lenses.

Depending on site lithology, direct-push delivery can be impractical or impossible due to rocks or subsurface debris. This type of environment would make injection well galleries a more suitable choice of delivery systems. It should be noted that plugging of the aquifer near the injection well might occur due to the formation of colloidal solids or precipitates. This could render an injection well useless for further injection and may create a preferential flow-path away from the target zone. If the injection well delivery method is selected, less costly well construction alternatives should be evaluated to avoid rendering expensive wells unusable. Flexible well-point designs are available from various vendors that could reduce down-hole material and construction costs significantly.

The need for further testing and evaluation

As with most tests, the MTBE bench scale analysis generated answers, as well as more questions.

• Is KMnO4 treatment effective at oxidizing TBA? What if other contaminants were present in higher quantities?
    
• What about different oxidants such as NaMnO4 or H2O2?
    
• Do oxidants oxidize contaminants preferentially?
   
• How long would it take to see the same reductions in contaminants in a soil matrix? What is the best way to test this given the inherent heterogeneity of soil?

The data for MTBE looks very promising in the above tests. A field scale evaluation of a TBA/benzene plume has been started.
 

Top