National Bureau of Standards Special Publication 556. Proceedings of WASTE OIL RECYCLING UTILIZING SOLVENT PRETREATMENT C. J. Thompson and M. L. Whisman Department of Energy Bartlesville Energy Research Center Bartlesville, Oklahoma 74003 Abstract A research program was initiated at the Bartlesville Energy Research Introduction As recently as 1971 there was little or no visible research, either in industry or government, being directed toward a more efficient utilization of used lubricating oils. This was true despite an obvious decline in the re-refining industry from nearly 150 re-refiners in 1960 to fewer than 40 in 1975. The combination of factors causing this decline is well known to this group. The decline in re-refining, coupled with the increasing shortages of petroleum, have resulted in the burning of about 45 percent of the used automotive crankcase drainings, with only a few percent being recycled and far too much being indiscriminately dumped. It is estimated that the greatest single pollutant in the Atlantic ocean is used lubes that are dumped into sewers and streams and eventually find their way to the ocean. Faced with a mission of conserving our natural resources and related environmental concerns, the Federal Government is in a unique position of providing unbiased research in the field of used oil reclamation that can be transferred to the public domain without proprietary encumbrances or conflicts--and such a program was initiated in 1971. Experimental The principal objective of this program of research was to develop new and improved technology for reclaiming used lubricating oil and to stimulate work by others in the field. The first step in this development was the evaluation of existing processes [1]. From information derived from patents and the scientific literature as well as personal contacts, our research staff reviewed many different approaches to reclaiming used automotive lubricating oil. Among these were the prevalent acid/clay process, caustic treatment, direct distillation/clay, propane extraction, treatment with fuming acid, and others. We were aware that quality lubricating oils could be and were being produced by the acid/clay process, but increasing environmental pressures related to acid sludge disposal led to an aggressive search for improved processes--preferably a process that would economically produce a quality oil and be more environmentally acceptable. I Underlined numbers in brackets indicate the literature references at the end of this paper. A hydrocarbon compositional study [2-7] of many feedstocks by our staff showed only minimal regional and seasonal variations in feedstock composition. This composition characterization included typical re-refinery feedstocks from widely divergent locations across the continental United States. The contaminants were measured and reported for these feedstocks, and, more important, the hydrocarbon composition of each was determined by a combination of chromatographic and mass spectral techniques. The remarkable similarity of the composition of these feedstocks strongly suggests that consistent product quality can be achieved from a re-refinery for any season or geographical locale. An early conclusion reached by our staff was the desirability and perhaps the necessity of a full vacuum distillation to restore used oil to original quality and provide fractions based upon boiling range and viscosity to broaden the spectrum and versatility of end products. Distillation of raw used crankcase oil proved to be impractical because of coking and fouling of processing equipment. Our investigations showed that cracking of unspent additives and other contaminants was evident at temperatures below 450° F. The resultant fouling required periodic shutdowns to clear lines and equipment of built up coke that is detrimental to heat transfer in processing equipment. To minimize fouling, a pretreatment that would substantially reduce coking precursors was needed. We have investigated dozens of pretreatment steps [8], and at this time we believe that a solvent system composed of one part isopropyl alcohol, two parts normal butyl alcohol, and one part methylethyl ketone represents a compromise between the most efficient solvent precipitation system and the most practical system. Three volumes of this solvent system are used with each volume of oil. We have found that combinations of alcohols are very effective but that solvent-to-oil ratios of 8 to 1 make such systems impractical. Propane is an effective pretreatment solvent, but the equipment required to maintain propane near its critical point makes propane less attractive than the three-component solvent system in readily available facilities. The system that we have developed for the reclamation of used automotive lubricating oil consists of the following steps: Step 1. Using distillation, the oil is first dehydrated to remove water and light ends. The light hydrocarbons are burned to recover their caloric content, and the water is treated for discharge into the sewer system. Step 2. The dehydrated oil is mixed with the appropriate solvents and sent to settlers. It is believed that centrifugation, while more complex and expensive, might, in some cases, replace the settlers, but it is possible to operate continuously with a number of settlers. The sludge is allowed to settle for an appropriate period, after which the oil-solvent mixture is drawn off and sent to a solvent stripper while the sludge is drawn from the bottom of the settler and also stripped of solvent using heat. The solvents can be reused, and solvent losses are estimated realistically at about 0.5 percent. Step 3. The solvent-stripped oil is sent to a fractionation tower where a fuel cut and several lube oil cuts are taken. Step 4. These fractions are then clay-contacted individually, or they are blended prior to clay contacting. The clay-contacting step generally requires higher temperatures and more clay than do virgin-derived lube stocks. This is attributed to traces of metals and oxygenates left after solvent precipitation and distillation. Typically, about 0.4 pound of an activated bleaching clay such as Filtrol 20 is used per gallon of oil, and the temperature is elevated to near 425° F with steam sparge to produce an oil with a color of about 1-1/2 and a bland odor. Clay is removed by means of a filter press, and the oil is subsequently reformulated with appropriate additives for the designated service. The process yields are typically 70 to 75 percent based upon the dry oil feedstock or near 90 percent on an oil-only basis. Since the sludges produced from both the solvent treatment and the distillation step are essentially neutral, they offer potential for use in roofing and road asphalt applications. Oil produced by this process, for which two patents have been allowed [9,10], has been bench-tested exhaustively and has also been submitted for engine sequence testing [11,12]. Two oils re-refined by technology just described and one commercially re-refined automotive lubricating oil were subjected to engine test sequences IIC, IIIC, VC, and L-38. The commercially re-refined automotive oil successfully passed the IIC, IIIC, VC, and L-38 engine-test evaluations required by automobile manufacturers to meet the standards established for service SE. One oil processed by BERC-developed technology successfully passed IIIC and VC sequence tests and, with the addition of 1 percent corrosion inhibitor, passed the IIC evaluation. The second BERC-produced oil successfully passed the IIIC test, and this same oil with additional corrosion inhibitor passed the IIC evaluation after an initial fail. The L-38 test was not run on BERC-produced oils, but a bearingcorrosion bench test that is reported to correlate well with the full-scale L-38 was run on one of the two oils with a successful pass. These engine tests are believed to be the first documentation in the United States of successful passing of engine sequence tests by re-refined lubricating oils to meet standards established for service SE. Further BERC-produced oils are satisfactorily performing in fleet tests in State of Iowa vehicles. This program is now in its 15th month, and at the end of 2 years a number of these vehicles will be disassembled and compared with like vehicles in similar service on commercially re-refined oil and on virgin oil. All test oils were performing satisfactorily at the 4,000-, 8,000-, and 10,000-mile drain intervals. Some data seem to indicate marginal superiority with regard to performance of the re-refined oils relative to new oil. Final comparisons among these oils will be based upon deposit and wear ratings to be made by an independent laboratory and on the accumulated analytical data from oil change samples. Not all of the waste oil recycling program at the BERC is devoted to the development of new technology. Some areas of more basic research are in progress, with still other areas to be undertaken as time, personnel, and funds permit. Currently, our staff is conducting a closed-loop composition study designed to investigate basestock compositional changes resulting from repeated use and processing of a motor oil. The original oil for this study was blended within our facility from virgin-derived basestocks and fully formulated with additives. The composition of this finished motor oil was studied while aliquots of the oil were charged to staff-operated automobiles and driven in normal engine service. The used oil from this fleet was subsequently collected and re-refined using the BERC technology. Aliquots of both the used oil and the re-refined oil were submitted to composition studies. At least one more cycle of use and re-refining is scheduled for this oil. We hope to define the compositional changes accompanying use and re-refining, as well as to determine the feasibility of repeated processing. We also are developing a test method for evaluating coking and fouling tendencies of feedstocks to re-refineries. We have had some very encouraging results from this study and believe that we can accurately estimate the coking tendencies of any re-refinery feedstock at any selected temperature up to the point of thermal cracking of the hydrocarbons. This analytical tool will be useful for evaluating pretreatment effectiveness to reduce coking and fouling precursors in heat exchangers and distillation apparatus. A predesign cost estimate for a 10-million gallon per year re-refinery, based upon the BERC technology, was made by the engineering firm of Richard J. Bigda & Associates [13]. The preliminary conclusions from this study include an estimated capital cost of $3 million for the plant. Waste oil purchased at $0.15 per gallon could be re-refined into blendable lube oils for $0.39 per gallon. Considering a selling price of $0.47 per gallon in bulk, the return on investment before taxes would be about 45 percent. Inflation plays havoc with numbers such as these, even before they are published, but relative costs and profits are probably still valid. Bigda and Associates determined that the operation would be very sensitive to the cost of the used oil, but plant capacity could be doubled or tripled with only a modest increase in capital investment cost. Currently we are directing the major thrust of our investigations toward engineering-type studies in pilot-scale operations that will aid in design of largescale plants incorporating the BERC technology. Included in these studies are dehydration operating parameters, optimum solvent-treatment conditions, solvent stripping techniques, and fractional vacuum distillation. Two recent economic studies have been made that involved comparisons of the BERC technology to the acid/clay reclaiming procedure. The first of these studies, made by Richard J. Bigda & Associates, indicated that capital investment would show a slight advantage for the acid/clay process, but total operating costs favor the BERC Solvent process. A significant conclusion of a more recent economic study was that the BERC method produces re-refined oil at a cost lower than that of the acid/clay process, and, as such, the BERC technology has a significant advantage over the acid/clay process both economically and ecologically. With regard to other new processes that have appeared in the past and recently in the patent literature, we and other investigators have evaluated many of these for technical viability. We have found that many proposed processes are technically unsound. Therefore, we caution those seeking new processes for converting used automotive lubricating oil into a high-quality product to assess technical viability carefully before evaluating economic factors. Actually, the BERC process is the only new process for which published data are available to support its technical viability. No other new patented process has ever established, through published engine tests or fleet tests, that high-quality automotive lubricating oil can be produced using such methodology. A member of the BERC staff recently visited France with Mr. Francois Audibert, the principal investigator of the French propane extraction method. Mr. Audibert reported in this interview that propane extraction cannot compete economically with acid/clay technology. We were able to also visit Viscolube in Milan, Italy, which is one of only two plants that use the French propane process. Mr. Schieppati, manager of the Italian plant, reported that propane treatment currently is very marginal economically. However, Mr. Schieppati believes that with new legislation to control the competition for used oil in Italy by those who would burn it will improve the profit picture. Similar steps in the United States to discourage burning and increase availability of feedstock are needed. Currently at BERC we are looking toward the selection of a partner from the private sector for the purposes of designing, constructing, and operating a 10-million gallon per year demonstration plant incorporating the BERC technology. Timing for startup of this plant at current funding levels is for January 1980; however, we believe that increased funding can advance the startup to April 1979. We hope to demonstrate both the technical and economic feasibility of the BERC process through the operation of a successful plant to bring about a timely transfer of our technology to private enterprise. If our plans succeed, we believe that our country will benefit from both the conservation of a valuable resource and from reduced environmental pollution. Conclusions A new and improved process for re-refining used automotive lubricating oil to original quality has been developed at the BERC. Engine and fleet tests have shown the process to be technically effective, and an independent engineering study indicates the new process has an economic advantage over conventional acid/clay treatment. Supporting research of a more basic nature is currently being conducted at the BERC, including oil composition studies designed to estimate compositional changes related to engine use, as well as re-refining by the BERC process. A test method is currently under development that appears to correlate with coking and fouling tendencies of re-refinery feedstocks. The method is believed to have application in the evaluation of particular feedstocks, as well as indicating the effectiveness of the pretreatment process in removal of coking and fouling precursors from used lubricating oil. Engineering studies are in progress to provide the basis for design, construction, and operation of a 10-million gallon per year demonstration plant to evaluate technical and economic viability of the BERC process in full-scale application. |