United States Environmental Protection Agency
Risk Reduction Engineering Laboratory
Cincinnati, OH 45268
Research and Development
EPA/600/SR-94/026
March 1994
Arun R. Gavaskar
Robert F. Olfenbuttel
Jody A. Jones
This study evaluated the product quality, waste reduction/pollution prevention, and economic aspects of three technologies for onsite solvent recovery:
The low-emission vapor degreaser is a fully enclosed alternative to conventional, open-top vapor degreasing. It was found to reduce air emissions by more than 99%, compared to a conventional vapor degreaser of the same production capacity.
Compared to disposal, the atmospheric and vacuum distillation units reduce operating costs significantly. The estimated payback period for these units was found to be less than 2 yr. The low-emission vapor degreaser reduced operating costs by reducing solvent losses and labor costs. The estimated payback for this unit was approximately 10 yr. The cost estimates were based on a full range of considerations including equipment, engineering, installation, operation, maintenance, and energy use. The estimates did not, however, include potential changes in liabilities or impacts due to regulations planned or in the process of being implemented.
This Project Summary was developed by EPA's Risk Reduction Engineering Laboratory, Cincinnati, OH, to announce key findings of the research project that is fully documented in a separate report of the same title (see Project Report ordering information at back).
This study, performed under the U.S. Environmental Protection Agency's (EPA's) Waste Reduction and Innovative Technology Evaluation (WRITE) Program, was a cooperative effort between EPA's Risk Reduction Engineering Laboratory (RREL) and the Washington Department of Ecology. The objective of the WRITE Program is to evaluate, in a typical workplace environment, examples of prototype or innovative commercial technologies that have potential for source reduction or recycling. The study evaluated three technologies for recovering and reusing waste solvent on site:
The two liquid-distillation units were tested at industrial sites that have purchased and are using the units. The atmospheric unit was tested on spent methyl ethyl ketone (MEK) at a site where MEK is used to clean the spray painting lines between colors. The recycled solvent was reused for the same purpose, with the residue shipped off as hazardous waste. The vacuum unit was tested on spent methylene chloride (MC) at a site that manufactures wires and cables. The MC is used for cold (immersion) cleaning of wires and cables to remove markings (ink).
Atmospheric distillation is the simplest technology available to recover liquid spent solvents. Units that can distill as little as 5 gal or as much as 55 gal/batch are available. Some units can be modified to operate under vacuum for higher-boiling solvents (>135ºC). Contaminant components with lower boiling points than the solvent or that form an azeotrope with the solvent cannot be separated (without fractionation) and may end up in the distillate. The unit used in this study (Figure 1 [provided in source document]) was Model LS-55D,[1] manufactured by Finish Thompson, Inc. The distillation residue, often a relatively small fraction of the spent solvent, is disposed of as hazardous waste.
[1] U.S. Government Printing Office: 1994 - 550-067/80201
The vacuum unit tested, Model 040 is manufactured by Mentec AG in Switzerland and supplied in the United States by Vaco-Solv Chicago, Inc. It is configured similar to a conventional vacuum distillation system except that the pump, in addition to drawing a vacuum, functions as a heat pump (Figure 2 [provided in source document]). No external heating or cooling is applied. The heat pump generates a vacuum for distillation and compresses vapors for condensation. Model 040 is suitable for solvents with boiling points up to 80ºC. Spent solvent is continuously sucked into the evaporator by a valve. The vacuum drawn generates vapors, which are sucked into the heat pump, compressed, and sent to the condenser. The temperature stabilizes automatically according to the specific solvent and the ambient air. The condenser surrounds the evaporator to allow heat exchange between the cool spent solvent and the warm condensing vapors.
The product quality objective for the two liquid-distillation units was to show that the recycled solvent was of sufficient quality for reuse. One 55-gal drum of spent solvent was processed each day through the batch and continuous units. For each unit, one drum of spent solvent was processed in 12 hr. The atmospheric unit left 16 gal of residue and the vacuum unit left 3 gal. The amount of residue left behind is a function of the application and not the distillation units. Samples of the spent and recycled solvents were analyzed by standard ASTM methods to determine the improvement in quality. Virgin solvent samples also were collected at each site and subjected to the same tests for comparison.
During the vacuum unit test, the "virgin" sample was found to be a sample of MC obtained by the site from a solvent recycling company. The "virgin" solvent specifications meet the requirements for the company's application, and it has been used satisfactorily at the site in the past. The vacuum unit was being operated at a faster rate than recommended by the manufacturer. Because the unit's built-in condenser-evaporator heat exchange was not sufficient for this rate, site personnel had attached an air-cooled condenser at the outlet to restrict vapor loss to 4 gal/55 gal of spent solvent. To prevent the release of this vapor into the work area, the vapor was led through a pipe to the roof of the facility and discharged per state regulations.
Table 1 [provided in source document]shows the characterization results for samples from the atmospheric and vacuum units. In appearance and color, the spent samples varied vastly from the clear recycled and virgin samples. All the measured parameters showed a significant improvement from spent to recycled samples but were not quite up to virgin grade. The water content increase in the recycled samples from the atmospheric unit was traced to a slight leakage from the water-cooled condenser that was worn out due to long use. Repairing the leak after the testing, site personnel reported that the problem had been corrected.
MEK purity of the recycled sample from the atmospheric unit substantially increased from 78% to 85%. The large decrease in nonvolatile matter during recycling accounts for most of this increase. Of the 15% impurity in the recycled sample, 5% is water as discussed above. The remaining 10% impurity probably is due to the codistilling out of paint thinner solvents (proprietary blends) present in the spent solvent. MC purity of the recycled solvent from the vacuum unit was 86%, comparing favorably with the "virgin" sample purity of 90%.
Some performance characteristics of MC (a halogenated solvent) also were evaluated. The pH of the water extract of the recycled solvent was fairly close to the "virgin" value of 7. The spent sample pH of 5 indicates the presence of potentially corrosive components. The corrosion test on steel and aluminum (ASTM D2251) yielded noticeable corrosion only in the case of the steel strip placed in the spent solvent sample. No such corrosion was evident due to the recycled solvent, indicating that recycling improved the quality.
Table 2 [provided in source document] shows the waste reduction achieved by the two distillation technologies at the respective sites. Through recycling, large volumes of spent solvent waste were reduced to small volumes of distillation residue, which is disposed of as RCRA hazardous waste. Both MEK and MC are hazardous chemicals listed on the Toxic Releases Inventory (TRI). These solvents also are on EPA's list of 17 chemicals targeted for 33% reduction by 1992 and 50% reduction by 1995.
The economic evaluation compares the costs of each new technology to conventional practice. Table 3 [provided in source document] shows the major operating costs associated with disposal and the atmospheric batch unit. For the unit, recycling saved $10,000/yr. The purchase price of the atmospheric batch unit is $12,995. A detailed calculation based on worksheets provided in the Facility Pollution Prevention Guide (EPA, 1992) indicated a payback period of less than 2 yr.
For the vacuum unit (Table 4 [provided in source document]), savings from recycling are $18,300/yr. An explosion-proof vacuum unit costs $23,500. The payback period for this unit also was less than 2 yr.
LEVD currently is used in Europe, where vapor degreasers are regulated as a point source. Previous studies (Battelle, 1992) on conventional open-top vapor degreasers have shown that a large part of the solvent (more than 90% in some cases) is lost through air emissions, which are considerable even though vapor degreasers are required to have primary cooling coils (tapwater cooled) and a certain freeboard height. Air emissions are mainly workload-related, caused either by dragout of solvent on the workload itself (and subsequent vaporization) or by disturbance in the air-vapor interface during entry and exit of the workload. Other sources are convection and diffusion during startup, operation, idling, shutdown, and, to a small extent, equipment leaks. Air emissions are a concern for metal finishers because many solvents used in vapor degreasing have been targeted by EPA in the 33/50 Program. Environmental and Occupational Safety and Health Administration (OSHA) regulations have become more stringent.
Pollution control devices available for conventional vapor degreasers include:
Table 5 [provided in source document] shows typical cleaning cycle stages.
Most residual solvent vapor in the cold air is adsorbed on the carbon. A photoionization detector (PID) probe verifies that the chamber air has less than 1 g/m3 of solvent and signals the air compressor to release the seal on the lid to end the cycle. If the chamber air has more than 1 g/m3 of solvent, the cycle loops back to the desorption stage. The entire cycle is programmed and requires no operator attention except to load and unload the workload. Only a very small amount of solvent exhausts at the end when the lid is opened. The LEVD also works as a distillation unit to clean the liquid solvent in the sump. During distillation, the unit is switched on without any workload in the chamber.
Testing was conducted on the LEVD using perchloroethylene (PCE) solvent. Test runs were conducted on machined steel parts with and without cutting oil on the parts. Total cycle times were recorded for all completed runs. Because the same batch of parts was used for each run, parts were either cold (ambient) or hot depending on the cooling time between runs. Adding oil to the parts did not greatly affect the total cycle time, but the workload mass did. In all the runs starting with parts dipped in cutting oil, the cleaned parts were visually examined for traces of oil or dirt contamination. No contamination was noticed on the parts from any of these runs.
The pollution prevention aspect of the LEVD was the main focus of this technology. The completely enclosed design of the working chamber allows the potential for air emissions only when the cleaning cycle is complete and the lid is opened. Any solvent vapor not evacuated from the chamber during condensation or adsorption releases to the atmosphere.
Table 6 [provided in source document] shows the total cycle times and emissions recorded from the LEVD by a flame ionization detector (FID) probe inserted (for this test) into the working chamber below the designated vapor level. FID measurements began during the adsorption stage and continued until after the lid was opened. A second FID probe (ambient), positioned outside the unit near the lid seal, took continuous measurements all around the unit during operation, with special emphasis around the lid to ensure leak-proof design. Ambient levels (3 to 4 ppm) in the indoor facility on the test days were consistent.
Figure 4 [provided in source document] shows how a typical LEVD cleaning cycle ends. The same pattern was evident in the other runs. Time zero corresponds to the start of measurements when the FID probe in the working chamber was activated.
Just before the adsorption cycle ended, the chamber FID read 52 ppm, well below the targeted 1 g/m3 (150 ppm of PCE). When the lid was retracted, the chamber air had full access to the ambient. At this point, the chamber concentration dropped sharply as the residual solvent vapor in the chamber dispersed. The ambient FID probe showed a corresponding increase (to 6 ppm). Both FID readings soon stabilized to facility ambient levels (3 to 4 ppm).
Later, as the basket of cleaned parts was raised out of the chamber, the second FID probe was thrust into the basket near the parts. No elevated readings above ambient were sensed, indicating that the parts were free of solvent. Thus, there is a very small air emission from the LEVD when the lid is opened. In all the test runs, the solvent concentration was well below the targeted 1 g/m3 (150 ppm PCE), so 1 g/m3 is an achievable concentration. The volume of the working chamber is 0.6 m3. Assuming that all the residual solvent vapor (1 g/m3 maximum) in the chamber is discharged to the ambient area, the typical air emission through the opened top is 0.6 g (0.00132 lb)/cycle or less. It takes 1 hr to clean 560 lb of oiled steel parts. Therefore, the air emission from this LEVD mode is 0.00132 lb of solvent/hr.
A typical conventional open-top vapor degreaser cleaning at a similar rate (560 lb of steel parts/hr) typically would emit 0.147 lb of solvent/ft2/hr (EPA, 1989), or 0.662 lb of solvent/hr from its 4.5-ft2 opening during continuous operation. Therefore, the LEVD reduces air emissions by more than 99% compared to air emissions from the typical conventional open-top vapor degreaser (i.e., with a 0.75 freeboard ratio, primary cooling coil, electric hoist, and no lip exhaust) used in this calculation.
The OSHA exposure limit for PCE is 25 ppm for an 8-hr time-weighted average (TWA). Personnel air sampling (in accordance with OSHA guidelines) was not conducted during this evaluation, but PCE levels measured with the ambient FID at all times during operation (3 to 4 ppm) and at the edge of the chamber opening for about 2.5 min when the lid is retracted completely (<6 ppm) (Figure 4 [provided in source document]) are well under the OSHA exposure limit. The pollution prevention potential of this unit is further enhanced by its ability to perform as a liquid solvent distillation system for cleaning the sump solvent; this capability was not a part of this evaluation. When pollution prevention is an objective, the LEVD also affords greater production flexibility because it has none of the idling losses between loads or downtime losses during shutdown of the conventional degreaser.
Table 7 [provided in source document] lists the LEVD's major operating costs and the operating costs for a conventional open-top vapor degreaser with similar production capacity. With a vendor-quoted purchase price for the LEVD of $210,000, the unit results in savings in annual total operating costs of $25,000 mainly from reduced labor costs (due to larger batch size) and lower solvent requirement (due to solvent recovery). The LEVD pays for itself in 10 yr. The above is a straightforward cost comparison between the LEVD and a conventional vapor degreaser of similar production capacity. Other cost-benefit factors must be taken into account when making economic decisions. The LEVD does not require capital and operating expenditures for auxiliary equipment that may be required for a standard conventional vapor degreaser (increased freeboard ratio, refrigerated coils, lip exhausts, room ventilation) in order to meet or anticipate increasingly stringent environmental and worker safety regulations. The LEVD is a self-contained unit that requires no additional facility modifications to achieve significant emission reductions.
Another consideration is the LEVD's production rate. The above calculation used a production rate of 560 lb/hr of steel parts (workload) because most vendors of conventional degreaser quote capacities based on steel parts. However, production capacity per machine can vary depending on the metal processed. Based on the thermal diffusivity of various metals, total cycle times versus production rates are plotted in Figure 5 [provided in source document]. Brass and copper can be processed faster than steel with the LEVD, and aluminum can be processed faster up to a point determined, for a certain shape of parts, by the maximum mass of aluminum parts that fit into the basket.
The shape of the parts also may affect cycle time. Parts with recesses that can trap solvent should be arranged in the basket so that the solvent liquid drains out. Other features offered by the vendor (oscillating or rotating baskets) should be used. Otherwise, either the air recirculation stage time must be increased, or the unit will loop into several adsorption cycles until the chamber concentration falls below 1 g/m3.
All three technologies evaluated in this study demonstrated good potential for pollution prevention/waste reduction. The two onsite solvent distillation technologies reduced large volumes of hazardous solvent to a few gallons of distillation residue and produced a reusable recycled product. The total U.S. solvent demand is approximately 160 billion gal/yr. Therefore, there is considerable potential for recycling and reusing spent solvent. Between onsite and offsite recovery, onsite recovery is preferable because of the reduced transportation hazard.
The largest single use for solvent in the United States is for vapor degreasing. The LEVD reduced air emissions significantly compared to emissions from a conventional vapor degreaser.
Payback periods for both distillation technologies are less than 2 yr. The LEVD is a slightly higher capital investment (with a payback period of approximately 10 yr), but it eliminates the need for other potentially expensive auxiliary equipment that conventional vapor degreasers would require to meet comparable pollution prevention objectives.
The full report was submitted in partial fulfillment of Contract Number 68-C0-0003, Work Assignment 2-36, by Battelle under the sponsorship of the U.S. Environmental Protection Agency.
Arun R. Gavaskar, Robert F. Olfenbuttel, and Jody A. Jones are with Battelle, Columbus, OH 43201.
Ivars Licis is the EPA Project Officer (see below).
The complete report, entitled "Onsite Solvent Recovery", (Order No. PB94-144508; Cost: $19.50, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
Last Updated: December 18, 1995