scope:

This report presents the results of a study sponsored by the American Petroleum Institute (API) to evaluate the effectiveness of five technologies which were used to treat petroleum refinery wastes. Technologies specifically excluded from this study were incineration, which has been tested extensively by the U.S. Environmental Protection Agency (EPA), and land treatment, which has been studied and reported on separately by API (API Publ. 4455).

Feed and product samples were collected at various test sites where waste treatment was applied. These samples were subjected to physical/chemical analyses to determine the effectiveness of various treatment processes.

This chapter (2) of the report describes the scope of the study, the technologies, the analytical scheme and the methods of data correlation and interpretation. The approach regarding the test site selection and the scale of treatment operation (commercial, pilot or laboratory), the analyses performed and the quality assurance/quality control methods used are also described in this chapter.

Chapters 3 through 7, provide information on each of the five treatment technologies tested. Chapter 8 provides a comparison of the effectiveness of all of the technologies tested in this study.

A schematic of the study, the treatment processes and the sequences of processes tested, are shown in Figure 2-1.

TECHNOLOGIES EVALUATED

Mechanical Treatment

Belt filters were tested at refineries C1 and C2, using listed oily wastes generated at those refineries. Operating conditions of each test were documented, Representative samples of waste feed, filtrates and filter cakes were transmitted to Rocky Mountain Analytical Laboratory (RMAL) for analysis according to EPA-prescribed protocols.

Plate filters were tested in a similar manner at refineries B1, B2, and B3. A rotary vacuum filter was tested at refinery V1.

Centrifuges are currently being used at a number of refineries but were not tested in this study for reasons of location plus the need to limit study costs.

Solvent Extraction

A solvent extraction process was tested which was thought to be representative of the general class of solvent extraction technologies. Tests were conducted on a 50-50 mixture of two listed wastes from refinery D. A batch pilot plant unit was used for this demonstration.

Thermal Treatment

A screw flight dryer was tested to represent this class of treatment technologies. Tests were conducted on belt filter cake from refinery C1, and on plate filter cake from refinery B1. Samples of each cake were treated at two temperatures--400°F (low temperature) and 650°F (high temperature) in batch at a vendor's pilot-scale facility.

Pyrolysis

A rotary pyrolysis process was tested--in this case by the process vendor following protocols established by the API task force--to represent this class of treatment technologies. The tested feed was a mixture of three listed wastes from refinery E.

Fixation

Three different fixation processes were tested on samples at three different levels of pretreatment. Untreated oily wastes (API separator bottoms and slop oil emulsion solids from refinery A) were treated with fixation process 1. Belt filter cake from refinery C1 and plate filter cake from refinery B1 were treated using all three fixation processes: 1, 2 and 3. Thermally dried belt and plate filter cakes were treated using fixation process 2.

DATA AND CORRELATIONS

Figure 2-2 summarizes the types of analytical data obtained from the technology testing and the correlations of the data that were used to assist in interpreting the test results. The general approach to data analyses is discussed by category below.

Analytical Data

As shown graphically in Figure 2-2, analytical data were of three types:

1. oil/water/solids analyses of the feed and products of each treatment technology;

2. analyses for Appendix VIII constituents in feed and products: and

3. Toxicity Characteristic Leaching Procedure (TCLP) analyses on feed and product solids.

More specific information on the analytical procedures used is given in a later section of this chapter.

Correlations

The analytical data listed above give a large measure of the effectiveness of the various treatment technologies: the concentration of Appendix VI11 constituents in the product solids, and the TCLP analyses on product solids. Two additional measures of treatment efficiency were developed: reduction of constituents by weight, and percent reduction in leachable concentrations. The first of these, Percent Reduction (Weight) is derived from a combination of the mass balance (for the most part calculated from oil/water solids analyses) and the constituent concentrations in feed and product solids. When combined, these data allowed the reduction in weight of constituents from feed to product solids to be calculated.

Each process was evaluated using this mass balance approach to determine the bulk amount of hydrocarbons or toxic Constituents physically removed from the waste and recycled. The main objective was to generate data which would provide a relative scale of the removal efficiencies of the generic processes, whether or not they could be considered viable BDAT technologies.

The wastes varied substantially in composition from very low to very high oil content. This was desirable because it gave a good range of wastes that would be expected to be processed through these units. However, this variation discourages absolute comparisons within a generic technology (e.g. plate versus belt filters) because some wastes were inherently easier to separate than others.

The second measure of treatment efficiency, Percent Reduction (Leachate Analysis), is derived from TCLP analyses on feed and product solids. The reduction in leachable concentrations of constituents from the feed to the product solids was determined.

LOCATION/SCALE/TECHNOLOGY SELECTION

A goal of this study was to test commercial-scale technologies, where possible, to demonstrate industrial application, availability and treatment performance. All tests on mechanical treatment equipment were conducted on commercial equipment operating at petroleum refineries. The solvent extraction process was tested on a pilot unit using a mixture of refinery sludges. A possible alternative choice would have been to test the commercial scale unit which was operating at a Superfund site. The decision was made that the testing on refinery wastes on a pilot scale was preferable to testing non-refinery wastes on a commercial scale since the two wastes had only a few constituents in common.

Thermal treatment was tested using refinery wastes on a pilot unit from an equipment vendor. Industrial operations other than petroleum use thermal driers on a commercial scale but no such operation exists treating refinery sludges. Feeds which were thermally treated were filter cakes. These were product solids derived from filtration of raw refinery wastes. Use of filter cake was based on a most probable scenario, but does not indicate any technical reason against feed of raw wastes to this equipment.

Pyrolysis and fixation tests were conducted at the laboratories of the process vendors. The process vendors were confident of their ability to scale up the tests based on experience with other feedstocks. For fixation, no advantage was seen in larger scale tests--all were at lab scale.

Selection had to be made among a wide array of process and equipment vendors based on a limited budget and time schedule. Results of the five technologies which were evaluated by API are likely to be indicative of generic classes or groups of processes and equipment.

ANALYTICAL PROCEDURES

The following sections give more information on the analytical tests and techniques and also summarize the quality assurance/quality control methods of the study.

Analytical Techniques

The oil/water/solids analysis employed a method developed by Chevron Corporation (Modified Oven Drying Technique or MOD-T.). The method is based on a low temperature distillation of the sample to generate a volatile oil and water fraction which is subsequently condensed. The remaining material is extracted with methylene chloride to generate a nonvolatile fraction, with solids defined as the solvent insoluble residue. This was the method of choice over EPA methods (i.e. 418.1, 3540 or 3550) specifically because of the potential for loss of volatile hydrocarbons by the EPA procedures. Consequently, results may differ among the test procedures, with the MOD-T reflecting somewhat higher oil recovery levels. A copy of this procedure is in the appendix.

Table 2-1 shows Appendix VIII constituents of refinery wastes which were analyzed in feeds, products and TCLP leachates. Analytical methods were derived from three sources of EPA methods:

1) the methods promulgated in 40 CFR 136 for priority pollutants;

2) the methods published in SW-846; and

3) methods published by EPA for Superfund investigations.

A subset of the Table 2-1 lists is shown in Table 2-2. This list was developed to allow a screening-test as an economy measure. These compounds termed "indicator" or "screening" compounds were measured-by alternative techniques.

TCLP leachates were prepared using the method in Appendix I to 40 CFR Part 264. Simply stated, the TCLP procedure is designed to generate an aqueous leachate of a waste. Leachates were prepared at a 20 to 1 ratio relative to the solid material in the sample. The leachate was then analyzed for the various target parameters. Results are reported in mg/L in the leachate.

The initial leaching procedure requires two separate laboratory preparations (extractions), one for volatile organics and one for the remaining parameters. The preparation for volatile organics requires the use of a specially designed device, termed the zero headspace extractor (ZHE).

The initial step in performing a TCLP extraction is the pressure filtration (50 psi) of the sample through a 0.8 micron filter. The solid phase remaining after this filtration is then mixed with the aqueous TCLP extraction fluid using a 20 to 1 ratio. After 18 hours of "extraction" the solid/leachate mixture is filtered a second time. The filtered leachate from this step is then combined with any filtrate from the initial filtration.

For wastes containing "oil" the initial filtration often results in a two-phase filtrate containing oil and water. According to the TCLP protocol, the oil phase must be analyzed separately, and the results mathematically combined with those from the extract. Analyses of the various leachate solutions were performed according to EPA procedures described in Appendix I (40 CFR Part 264).

Samples of fixed waste materials were ground to pass a 5.55 mm (0.375-inch) standard sieve prior to conducting either the Total TCLP or the Indicator TCLP test as required by the EPA protocol (51 FR 40643).

Quality Assurance/Quality Control

All laboratory analyses were performed according to specifications in a Quality Assurance Project Plan (QAPP), as specified in EPAAR 1552.246-71. The QA/QC plan of the laboratory (RMAL) for this project followed the elements of their generic laboratory-wide quality insurance procedures for sample preparation and analyses. A separate QC Data Summary Report has been prepared by RMAL, which presented the QC results that were directly related to the performance of the methods on these samples. Other QC activities such as calibration, mass tuning checks and activities related to the general performance of the instruments have not been reported, but are archived in the report files at RMAL.

Quality control analyses consist of the following activities which are included in the QC Data Summary Report:

• multipoint standard calibration;

• analysis of blanks;

• analysis of spiked and duplicate samples;

• analysis of standard reference materials;

• daily calibration, including mass spectrometer tuning checks (BFB and DFTPP), where appropriate; and

• addition of surrogate spikes into each sample for GC/MS analyses.

Sufficient amounts of representative samples were sent to the laboratory with the following history:

• All samples were collected in glass sample bottles:

• No chemical preservation was used, and the samples were stored at 4°C until analyzed;

• Samples for volatile organics were collected with minimal headspace; and

• Analysis was performed in an expeditious fashion, applying 40 CFR 136 water holding times where appropriate.

A chain of custody record was established for each sample except as noted in the QA/QC report.