Evaluation of land farming and chemico-biological stabilization for treatment of heavily contaminated sediments in a tropical environment

Abstract

*R. H. Adams; F. J. Guzmán-Osorio

Laboratorio de Biorremediación, División Académica de Ciencias Biológicas, Universidad Juárez Autónoma de Tabasco. Carretera Villahermosa-Cárdenas Km 0.5, Villahermosa, Tabasco, México

Conventional and experimental methods were studied for the remediation of petroleum contaminated sediments from a dam previously used to collect acid run-off from a sulfur mine. The man-made lake had been neutralized, but bentonite rich sediments remained contaminated with very weathered hydrocarbons (sediments with ~50,000–60,000 mg/kg Total Petroleum Hydrocarbons were used in this study). Biostimulation, bioaugmentation (with native microorganisms) and chemico-biological stabilization, all resulted in similar reductions (14–16%) in the TPH concentration over a three month period. The land farming treatments resulted in variable reductions in toxicity, ranging from nil to complete, while the chemico-biological stabilization treatment, not only eliminated acute toxicity but also resulted in a slight stimulation (~103–109%) of the test organism in the bioassay (Microtox). All three treatments reduced polyaromatic hydrocarbons of probable carcinogenicity to below or nearly below the Mexican norms, reduced Toxic Characteristic Leaching Proceedure leachates to <1 mg/L, and left the material in a pH range of 7.0–7.8. The chemico-biological stabilization has the advantage of only requiring initial mixing of the chemical and organic reagents instead of daily aeration, thereby reducing operating costs. This method is also able to treat very difficult sites at low cost, relying on biological humification processes which are accelerated in a humid tropical and semitropical environment. The total unit cost of the chemico-biological stabilization treatment was estimated to be ~60% of that for land farming in the southern Gulf of Mexico region.

Key words: Bioremediation, oil, total petroleum hydrocarbons, soil, toxicity

An open access article from Int. J. Environ. Sci. Tech., 5 (2), 169-178, Spring 2008, viewed from Biology-Online.org.

Introduction

Most of the methods currently used to remediate contaminated soils in emerging countries are mere adaptations of technologies originally developed in industrialized regions. However, the imported methods were designed for areas with different kinds of economies, cultural expectations and frequently, different physical and biological environments. The misuse of the imported technologies is often inadequate to the local needs when employed in developing regions. Furthermore, their implementation frequently requires the importation of materials, machinery and know-how. This creates dependence on foreign technology and increases the cost of remediation. Unfortunately, the imported technologies are usually poorly suited to the new socio-economic and physico-biological environments, and fail to succeed (McMillen et al., 2002). Besides, in many developing regions, especially in humid tropical and subtropical environments, physical, chemical or biological processes occur which can be applied to the remediation of contaminated soils (Adams and Castillo, 2000).

An exception to this tendency is the Chemico- Biological Stabilization process for treating hydrocarbon contaminated soils, sediments and drilling cuttings. This process was developed in the southern Gulf of Mexico region and takes advantage of the increased rate of natural biological processes in a humid tropical environment. This method has been proven to be able to treat hydrocarbon contaminated soil and drilling cuttings with reduced labor, machinery and fuel costs rather than other commonly used methods. In the present article, this technology is compared with biostimulation (land farm) and bioaugmentation considering effectiveness, logistics and costs.

This method was developed by and for people and sites in the southern Gulf of Mexico region and can be implemented using locally available materials, machinery and know-how. The first steps in developing this method came from petroleum industry personnel trying to treat old drilling waste and purge pits using simple procedures that had been observed to reduce the oily appearance of contaminated materials, namely the addition on hydrated lime (Ca(OH)2) or quicklime (CaO) plus an organic absorbent (sawdust). These as well as organic rich river levee soil were mixed with the drilling and/or purge wastes and then extended as a thin layer (approx. 20–40 cm) over the surface of the site, and later planted with local grasses or allowed to become naturally colonized with local weedy vegetation. Originally, petroleum industry personnel considered adding the river levee soil to be an important source of microorganisms for biodegradation of the hydrocarbons. Using this method, some sites were partially treated, but the national environmental agency responsible to inspect remediated sites (the Federal Attorney for Environmental Protection, PROFEPA for its abbreviation in Spanish) considered that, due to the large additions of other materials, the process was simply a dilution (at that time, the addition of amendments and soil used by petroleum personnel were quite large). These sites faced serious difficulties in being considered as remediated by the authorities (PEMEX, 1995; Arroyo, 1997).

Subsequently, this method was developed technologically in a local state-run university allround the region. The method was tested systematically, altering: 1) the amount of chemical reagent, 2) the amount of organic amendment, 3) the type of organic amendment, 4) the order of application of chemical and organic amendments and 5) the elimination of exogenous soil (Adams, 2004a). It was found that the order of application was very important and the addition of organic rich soil was unnecessary (Adams, 2004b). Also, other organic amendments worked much better than the sawdust originally employed and important advances were achieved over a mediumterm period with respect to soil fertility (field capacity and water repellency, Adams et al., 2007). Furthermore, toxicity and soil leachates were demonstrated to be considerably reduced in the short term and not to present an unacceptable risk to soil or groundwater (UJAT, 2006).

Material was collected from a sediment “beach” area in a large (320 Ha) dam used to temporarily store waste runoff from a sulfur mine in the southern Veracruz State (Mexico). Sulfur was mined at this site from salt domes using the Frasch process, and acid leachate from the tailings pile was diverted to the holding dam (UNAM, 2002). During the drilling process, bentonite-based drilling fluids were used and at least part of the spent muds were also diverted to the dam, forming a large (approx. 23 ha) sediment “delta” or beach area in which the fine clay settled out (UJAT, 2006). Interlaced with the fine sediments are thin veins of very viscous hydrocarbons which originated from the salt domes and were considered waste material by the mine operators and discharged with the acid water and drilling fluids. This oil is slightly denser than water and has settled on the bottom of the dam just above the sediments. Outside the sediment delta area, there has been very little infiltration of this oil into the native sediments due to its viscosity. Only 13% of the dam area (not including the sediment delta) has hydrocarbons in the native sediments although approx. 60% of this same surface is covered with a layer of oil (UJAT, 2006). The very high viscosity of the oil has prevented further penetration. In gas chromatography characterization (EPA, 1997), the oil was found to be generally midrange (10 - 25 C) but extremely weathered without any peaks arising above the undefined complex mixture in chromatograms. The analytical laboratory, at which samples were processed, sent the following footnote on the analytical report: “appears to be weathered lubricant oil”. The concentration of hydrocarbons in the sediments used in this study was approx. 5–6% (dry weight). The water in the dam had been neutralized with magnesium oxide as part of the site remediation prior to lowering the level in the dam and collecting the sediments used in this study.


Materials and Methods

Sediment collection

Material was obtained from the sediment delta at the dam “Presa Agua de Minas”, in the Texistepec Mining Unit in Texistepec Municipality, in southern Veracruz State, Mexico, in March 2004. In the area where material was collected, an oily crust of approx. 10 cm had already been removed manually as part of the remediation of the site. Material was collected with a straight shovel from the surface to a depth of 90–100 cm, obtaining roughly one-half cubic meter of the material. This material was placed in hard plastic containers and transported to the remediation patio of the Bioremediation Laboratory in the Juarez Autonomous University of Tabasco (Universidad Juárez Autónoma de Tabasco, UJAT), where it was thoroughly mixed.

Experimental design

The material was treated by three methods:
1) bioremediation – land farm with biostimulation only,
2) bioremediation – land farm with bioaugmentation,
3) chemico-biological stabilization, and a control (without treatment). Each treatment and control included three replicates. For each replicate, the material (after being mixed) was placed in cells 40 cm x 40 cm x 20 cm depth. The cells were formed of cinderblock and situated on a concrete slab. All tests were run in the open air in a tropical environment: ~ 1800 mm annual average precipitation, ~ 29 ºC average annual temperature (West et al., 1987). The three treatments are described below:

Biostimulation: addition of inorganic nutrients (Grow-Feed 20-30-10 commercial fertilizer) sufficient to provide 100 ppm of nitrogen, an organic amendment (0.74% of chopped water lily, dry weight basis), daily aeration and periodic moisture addition (to maintain approx. 15 -22% moisture). The actual moisture maintained in the treatment cells was 19.3% ± 1.51%. Temperature was measured daily and overly high temperature was controlled by spraying a small quantity of water on the treatment cells. The average temperature in the cells during the study was 32 ºC ± 2.7 ºC.

Bioaugmentation: fertilizer and organic amendment addition as in the biostimulation treatment, followed by the addition of autochthonous microorganisms, aeration, and moisture control (to maintain approx. 15- 22% moisture). Moisture and temperature were controlled as described for biostimulation, with the actual moisture maintained at 19.4% ± 2.21%, and the average temperature was the same as that in the biostimulation treatment (32 ºC ± 2.7 ºC). Chemico-biological stabilization: addition of 4% (w/ w) hydrated lime, mixing, followed by adding 4% (w/w) organic amendment (cachasse: sugar cane process waste) three days latter (Adams, 2004b). Fifteen days later, humidicola grass (Brachiaria humidicola) seeds were planted on the surface of the treatment cells.

Preparation of autochthonous microorganisms for bioaugmentation

Contaminated soil from the sediment delta was used as a source of microorganisms to prepare an undefined mixed culture. A 1/10 dilution of field moist sediment in deionized water was prepared, mixed thoroughly, and allowed to settle for 20 min. One milliliter of the supernatant was collected, taking care to avoid floating particles, and used to inoculate 100 mL of mineral nutrient which consisted of 1 g/L of soluble commercial fertilizer (Grow-Feed 20-30-10) in deionized water. As the C source, a drop of oil recovered from the contaminated sediments was added to the culture flask on the first day of incubation, followed by a drop of spent lubricant (motor oil) on the second day (the oil from the contaminated sediment was extremely viscous and produced a relatively low level of growth by itself). The culture flasks was placed on a New Brunswick Scientific model 01071-0000 orbital shaker (~180 RPM) to aerate the culture, at room temperature (~28 ºC). After three days, a notable increase in turbidity was observed, and the culture was used to inoculate a second culture flask using 1 mL of the first culture to inoculate 100 mL in the second culture. This process was repeated (three sequential cultures in all) to ensure that the selected microorganisms were actually using the added hydrocarbons as the C source and not alternative C sources from the original dilution of the sediment. In the third sequential culture, growth was measured with a Spectronic 20 (Bausch and Lomb) spectophotometer at 600 nm using uninoculated growth medium as a blank. This third culture was started using oil from the site as a C source and after roughly three days a drop of spent lubricant was added. After another three days, another drop of spent lubricant was added. After 12 days, this culture showed an increase in absorbance of > 0.2 O.D. and was used to inoculate the final culture used for bioaugmentation in the treatment cells.

One and one-quarter milliliter of the third sequential culture was added to each of twelve 250 mL flasks, each containing 125 mL of mineral growth medium (1 g/L of Grow Feed fertilizer). One drop of oil from the site was added to each flask to start the culture. After three days and again after another three days, one drop of spent lubricant was added to each flask. After 12 days, the cultures were harvested, pooled and divided into three 500 mL portions.

Each portion was added to a bioaugmentation treatment cell, applying the culture with a manual spray bottle, mixing the solution into the soil with a garden shovel. The final inoculation was estimated to be ~ 6 x 106 bacteria/g of dry soil considering the turbidity of the final culture and the application rate.

Monitoring

Initially and on a monthly basis, core samples were taken from each treatment cell, from the surface to 20 cm depth and analyzed for a variety of parameters. pH was measured with a Corning brand, model 240 pH meter with a Sensorex brand, model 5200c pH electrode (SECOFI, 1980).

Total Petroleum Hydrocarbons (TPH) were measured by EPA method 418.1 (EPA, 1997), using perchloroethylene as an extraction solvent with an Infracal TOG/TPH analyzer (Wilks Enterprise). Acute toxicity was measured with the Microtox system using a Microbics model 500 analyzer (SECOFI, 1996). Toxicity Characteristic Leaching Proceedure (TCLP) leachates were prepared (SEDESOL, 1993) and TPH in the TCLP leachates was measured by extraction in perchloroethylene, solvent evaporation, re-dissolution in methanol, and measurement of hydrocarbons in the methanol solution using EPA method 9074 (EPA, 1997; Adams and Ramírez, 1999). For this TPH analysis in leachate a Petroflag soil analyzer (Dexsil Corporation) was used. Samples for carcinogenic Polyaromatic Hydrocarbon (PAH) analyses were sent to a commercial laboratory accredited by the Mexican Accreditation Entity (EMA for its abbreviation in Spanish) and were analyzed by EPA method 8270 D (EPA, 1997) by gas chromatography/mass spectrometry (GC/MS) using a Hewlit Packard HP5890 GS coupled to a Hewlit Packard HP5971 MS detector, or equivalent.


Results and Discussion

Changes in hydrocarbon concentration

In Figure 1, the results of monitoring for hydrocarbon concentration are presented. All data presented are for core samples from 0-20 cm depth. During the first month of treatment, all of the processes registered a reduction in TPH concentration: ~20% for biostimulation and bioaugmentation and 12% for the chemico-biological stabilization. However, during the second month of treatment, an increase of 4–9% in TPH concentration was observed. It is possible that this observation is due to changes in the interaction between the hydrocarbons and soil particles that make the hydrocarbons more available and results in an increase in the extraction efficiency. If the soil microorganisms begin to degrade some of the organic material in the soil (humic substrates, water lily or cachasse), this could liberate part of the hydrocarbons from soil particles. An alternative explanation could be that biosurfactants are produced by oil-degrading microorganisms, also leading to greater availability and increased extraction efficiency. It is also possible that some of the intermediates of biodegradation (alcohols, for example) produced greater IR absorbance, or were more easily extracted from the soil matrix than the original hydrocarbons. Increases of this type have also been reported by other authors during the intermediate phase of biodegradation (Adams et al., 2008; Bartha and El-Din, 1993).

During the third month of treatment, this trend (increased TPH concentration) continued in the bioaugmentation treatment (increase of an additional 5%), but in both the biostimulation and chemicobiological stabilization treatments, important reductions were observed to be about 3.5% for biostimulation and 9% for the chemico-biological stabilization treatment. Overall reduction during the three month period was ~ 14–16% in each of the three treatments. Interestingly, the gas chromatography of oil from this site showed 14% of the oil to be in the C10-C15 range (with 86% C>15). This overall low rate of biodegradation of very weathered hydrocarbons in soil is not unusual (Adams, et al., 2006; Atlas, 1986; Jerger et al., 1991). It probably results not only from more complexity in their chemical structure, but due to their high viscosity and reduced bioavailabilty. Similar kinds of hydrocarbons in liquid broth or slurry conditions (where surface area and bioavailability problems are mechanically overcome) have been highly biodegraded in relatively short time periods (Pradhan et al., 1997; Jerger et al., 1991).

Reduction in toxicity

The results of the toxicity evaluation are shown in Figure 2 In this scale, toxicity is represented in toxicity units=1/EC50, where EC50 is the effective concentration-50, the concentration of the sample which reduces the activity (bioluminescence) of the test organism in the bioassay by 50% (Bulinc, 1990). In this case, the EC50 is calculated as a proportion (75 % = 0.75, for example). The scale used in this study is the quarter log scale proposed by Kross and Cherryholmes (1993) normalized to the background toxicity at the site (Cornelio, 2001; UJAT, 2006). Samples in which no toxicity patter