Ames, Iowa

REPORT WRITTEN BY:
John M. Yacher
William A. Heitbrink
G. Edward Burroughs

REPORT DATE:
July 25, 1997

REPORT NO.:
ECTB 218-12a



U.S. DEPARTMENT OF HEaltH AND HUMAN SERVICES
Public Health Service
Centers for Disease Control and Prevention
National Institute for Occupational Safety and Health
Division of Physical Sciences and Engineering
Engineering Control Technology Branch
4676 Columbia Parkway, Mail Stop R5
Cincinnati, Ohio 45226-1998

ABSTRACT

This study evaluated the ability of commercially available air filtering cleaners installed on more than 25 machining centers to control mist emissions and to reduce workers' mist exposure. In a machining center used to produce transmission parts, a mist of synthetic metalworking fluid (MWF) was generated as a result of drilling and tapping holes at rotational speeds of 1000 to 3000 rpm. The MWF was flooded over the parts at 80 psi during most machining operations. To facilitate metal chip removal during some operations, MWF was pumped through the orifices in some tools at a pressure of 800 psi. These machining operations were performed in nearly complete enclosures that were exhausted to air cleaners, whose fans moved approximately 2400 cfm of air.

To evaluate air cleaner performance, the concentration of triethanolamine and total particulate were measured before and after the installation of the air cleaning units. Area concentrations were reduced from a high of 0.48 mg/m³ to 0.04 mg/m³ or less. The total particulate concentrations on the personal samples showed a four-fold decrease from 0.22 mg/m³ to 0.06 mg/m³.

An aerosol photometer (HAM, ppm, Inc., Knoxville, TN) and video monitoring were used to identify peak exposures to machine operators in the course of their work. Peaks occurred when operators entered or partially entered the machining center enclosures. Some sources of increased air contamination were identified by use of an eight-channel optical particle counter (Portable Dust Monitor, Model 1105, Grimm Ainring, Germany); the most significant sources were partially or unenclosed machining centers and inadequately covered flumes returning the MWF to the Hydromation (fluid recirculation and filtration) unit.

A quartz crystal microbalance cascade impactor (model PC2, California Measurements, Sierra Madre, CA), and eight-stage particle fractionating samplers (1 ACFM Ambient Particle Sizing Sampler, Anderson Samplers, Inc., Atlanta, GA) showed that particles larger than 9 µm were present in the plant environment. This suggests that besides the machining centers tested, there were other relatively minor sources of particulate such as uncontrolled machining operations.

INTRODUCTION

Sauer-Sundstrand Company is a metalworking plant located in Ames, Iowa. In this location, there are approximately 300 employees in the production area and approximately 200 employees in the office area. Sauer-Sundstrand continues production 24 hours a day, with most production area employees working a 10 hour shift and a 40 hour week. Transmissions are produced for off the road vehicles such as lawn mowers and agricultural equipment. The iron castings which are brought in the plant are pre-shaped for the transmission. Additional metalworking is performed on the piece, including milling and drilling. Each metalworking station is automated. One operator programs and tends several machines.

Metalworking fluid (MWF) is also referred to as coolant, and the two terms will be used interchangeably throughout the text. It is used during the metalworking to remove metal shavings and to serve as a coolant and lubricant. At the metalworking stations examined in this study, the MWF was flooded onto the parts at a pressure of 80 pounds per square inch (psi). During some machining operations, the coolant is forced through small holes in the drills at higher pressures ranging between 600 to 850 psi. The high pressure application of fluid was used during approximately 30 percent of the machining cycle. In other machines, other coolant applications may reach pressures as high as 1200 psi. During the high pressure application of coolant, the tooling rotations reached as high as 4500 rpm, with an average of approximately 1000 rpm. The lower pressure applications flooded the part with the fluid at relatively low pressures, around 80 psi, approximately 70 percent of the machining cycle. The bottom of the machining center has a sloped bottom where the excess fluid and debris are removed via a screw feeder leading to the fluid recycle system. In the L-shop, the area studied during this survey, fluid is recycled through the Hydromation unit, which is used to pump and filter the fluid, removing metal chips and other debris. The Hydromation unit storage pit has a volume of 10,000 gallons. The fluid used in the L-shop at approximately 12 stations was Syntilo® 9902 (Castro Industrial, Inc., Downers Grove, IL), a synthetic product primarily composed of water and triethanolamine. Several different types of MWF are used throughout the plant at approximately 250 metalworking stations.

The main focus of this study was L-shop where it was thought that the majority of plant metalworking fluid mists were generated. In L-shop, metalworking was performed on items with a low volume total to be produced. High quantity orders were done elsewhere in the plant. There were 12 stations in L-shop with approximately 45 employees. These machining units were all partially enclosed and automated.

Study Objectives

Sauer-Sundstrand Company requested that NIOSH researchers perform an evaluation on the efficacy of a commercially available air cleaner. This air cleaner would be placed downstream from a metalworking station, and the "cleaned" air would be recirculated into the plant, thus saving heating and cooling costs. The recirculation would also eliminate the need for an exhaust stack for the numerous stations; also, there are significant time-delays associated with obtaining stack permits from local air pollution control agencies. In order to meet production demands and to save money, the air is circulated through an air cleaner and the discharged air is recycled to the plant. Thus, there is a need to evaluate the efficacy of air cleaners for removing MWF mists. Sauer-Sundstrand hoped to gather additional information in order to decide if this type of air cleaner should be installed throughout the plant on each of the metalworking stations. One of NIOSH's goals for conducting this in-plant study was as a prelude to future pilot plant studies to evaluate the effect of machining parameters upon size dependent mist concentrations. The main issues to be examined included the following:

Health Effects

There are many health effects associated with metalworking exposures including dermatitis,² respiratory disease,³ and asthma.⁴ Cross-shift decrements in lung function are reported for inhalable aerosol exposures larger than 0.2 mg/m³.³ Microbial contamination and endotoxins (debris horn dead microbes) may also be responsible for adverse pulmonary health effects? Some on-going research has suggested that lifetime exposures to specific types of metalworking fluids (straight, soluble, and synthetic) are associated with several digestive cancers.⁵ For these reasons, it is prudent to control exposures to metalworking fluids.

Exposure Evaluation Criteria

Triethanolamine is the major component of the synthetic MWF used during this study. For triethanolamine, the American Conference of Governmental Industrial Hygienists (ACGIH) has established a Threshold Limit Value (TLV) of 5 mg/m³ as an 8-hour time weighted average.⁶ The ACGIH is a private organization and its TLVs refer to airborne concentrations to which nearly all workers may be repeatedly exposed without experiencing adverse health effects.

The Occupational Safety and Health Administration (OSHA) has established a permissible exposure limit for particulate not otherwise regulated of 15 mg/m³ as an 8-hour time weighted average.⁷

Air Cleaner Description

The air cleaner installed is shown in Figure 1. It is a packaged air filter unit, Model F120, manufactured by Airflow Systems, Inc. (Dallas, TX) with an approximate cost of $4000. The units were installed over the metalworking stations and pulled the air into the cleaning units. The air cleaner's fan moved approximately 2400 cfm through the enclosure. The air cleaner is equipped with a metal mesh prefilter, followed by a pleated "mist eliminator" prefilter. Next are the main filters, which are 95 percent efficient ASHRAE pocket filters. According to ASHRAE guidelines, a 95 percent efficiency filter removes all particles with a diameter of approximately 2 µm. The fractional efficiency curve of this filter also shows a minimal efficiency of approximately 72 percent for particles sized near 0.3 µm.⁸ The fluids captured by the filters drip to the floor of the cleaner and exit via three drainage holes. The coolant then drains to the Hydromation recycling system. At the outlet of the cleaner, is a 4-way adjustable grill for the exiting air.

Figure 1: Installed air cleaner with incline and larger diameter drains; modifications made following qualitative observations of MWF accumulation noted during the second phase of the study.

Modifications to the Air Cleaner After Phase 2

During Phase 2 of the study, it was noted that the three air cleaner drains were clogging, perhaps entraining additional MWF mist into the air flow. As a result, the facility maintenance personnel enlarged the drains from 0.5 inches to 1 inch in diameter to allow proper drainage. Plastic translucent tubing was added to the drains, leading to a goose neck fitting which led, ultimately, to the Hydromation unit. The translucent tubing showed if there was fluid draining and would indicate if there was blockage. Also, the air cleaner was tilted slightly so that it sloped toward the drains instead of the fire. Maintenance personnel reported that the exhaust grill remained clean.

Air Cleaner Maintenance

A maintenance program was established for the installed air cleaners. Each air cleaner was fitted with an aneroid pressure gauge to indicate pressure drop across the system. It is not permitted to exceed 2 inches w.g. The filters are changed at 30 day intervals and thoroughly cleaned or replaced. Because the MWF tends to collect in the filters and not fully drain when a machining center is run on a 24 hour per day cycle, the air cleaners on these machines are turned off, i.e., rested for 1 hour out of each 24 hours. This "rest cycle" is not necessary for the machining centers that are run for 8- to 10-hour periods each 24 hours. With the exception of the "rest cycle," the maintenance program appears to be conservative and may help to explain the low concentrations measured.

Instrumentation

The instruments below were used during sampling:

Table 1:
Theoretical and Experimental Values of D₅₀ for Unit Density Spheres
for California Instruments QCM Obtained by Fairchild and Wheat⁹

EXPERIMENTAL PROCEDURES - PHASE 3

Impinger samples were taken using NIOSH Method 3509 for triethanolamine (TEA).¹⁰ Pump flow rates were increased to 2 lpm. The limit of detection (LOD) was 9 µg/sample and the limit of quantitation (LOQ) was 29 µg/sample. There were no samples reported as nondetectable. Samples which were found to be between the LOD and LOQ were estimated with the analytical laboratory reported result.

Impinger sampling was only used for area samples. Samples were taken in several areas, including "between machines" for 4-8 hour periods. Two other sites were chosen because they were at opposite boundaries of the L-shop; the samples were denoted as "L-Shop Edge" and near the Hydromation unit (Hydro), "central cleaning."

Personal and area samples for total weight particulate were taken according to NIOSH Method 0500.¹⁰ Area samples were taken at the same locations as for impinger samples. In addition, samples were collected on the workers in the area. Other than blanks, there was only one sample which resulted in a nondetectable level. For statistical purposes, this sample was estimated to be LOD/2, or 0.01 mg/sample.

Two operators were monitored by an aerosol photometer, handheld aerosol monitor (HAM) manufactured by ppm, Inc. The unit was belt mounted and the operators were videotaped as they performed various tasks including machining center adjustment and cleaning, re-mounting castings on "tombstones," checking specifications of machined castings inside a plant floor enclosed room, and making adjustments on partially enclosed machining centers. See Figure 2 for the plant floor layout and approximate locations of peak measurements.

The Portable Dust Monitor (PDM), an optical particle counter, manufactured by Grimm was used at six locations inside the plant shop areas (see Figure 2), the front office area on the first floor, and outside the main employee entrance. Total particulate concentration was reported.

The quartz crystal microbalance cascade impactor (QCM) manufactured by California Instruments was used in severaI locations to determine the particle size distribution in the plant atmosphere. Measurements were made near the "central cleaning" (Hydro) unit and "between machining centers."

Two eight-stage inertial impactors (Anderson), with no preseparators, coupled with one CFM pumps were used to collect size distributed total particulate in the same two plant locations sampled with the QCM, the impingers, and the filter cassettes. Total particulate was collected continuously for nearly three full days, including the one hour per day during the night shift that the air cleaners were turned off to allow the filter elements to gravity drain.

RESULTS AND FINDINGS

TEA and total particulate concentrations measured during Phase 3 are listed in Appendices I and II. The impinger data for TEA, comparing concentrations measured before and after controls (air cleaners) were installed, are summarized in Table 2 and Figure 3. Similarly, the filter cassette data for total particulate are summarized in Table 3 and Figure 4. Inspection of these tables and figures shows decreases in concentrations by factors of 2 to 10. The right most columns in the tables present the probabilities that chance could have caused the differences. With the exception of the total particulate concentration measured at the "L-Shop Edge" (Table 3), these differences are all very significant. Installation of controls reduced workers' personal total particulate exposure from 0.22 mg/m³ to 0.06 mg/m³; however, this reduced value was higher than area concentrations measured at the "L-Shop Edge" and "between machines," based upon the Wailer-Duncan k-ratio t test.¹² Thus the worker engages in activities which provide some minor increase in his total particulate exposure.



Figure 2: L-Shop floor plan showing sampling locations. B is between machining centers, C is the central cleaning (Hydro) unit, F is in the flume, L is L-Shop edge, P is a partially covered flume, and T is an unventilated machining center. The circled numbers indicate the video monitoring locations; 1 is for Operator A, and 2 is for Operator B.

Table 2:
Summary Statistics for Triethanolamine Concentrations

  Table 3:
Summary Statistics for Total Particulate Concentrations

Figure 3: Triethanolamine (TEA) concentrations measured before (August 1995) and after (August 1996) installation of air cleaners in the Sauer-Sundstrand plant.





Figure 4: Total particulate concentrations measured before (August 1995) and after (August 1996) installation of air cleaners in the Sauer-Sundstrand plant.

The results of the HAM/video-exposure monitoring are shown in Figure 5. Operator A had his highest reading, 0.93 mg/m³, when he was inside a machining center (#4897) "cleaning;" his second highest, 0.46 mg/m³, occurred when he was inside the L-shop floor office checking specifications (tolerances). Operator B had his highest levels when he was at the open door, at times with his arm inside, partially enclosed machining centers (#6922 and #6921). The levels were 0.45 to 0.63 mg/m³.

The results of the PDM (Grimm) survey are shown in Figure 6. The highest relative mist concentration levels were found in the flume near a machining center across the main isle from the Hydromation (central cleaning) unit at 22.4 mg/m³; the lowest, outside the plant at 0.018 mg/m³. Levels at an unventilated, partially enclosed machining center (#6902) were the next highest at 0.397 mg/m³ followed closely by 0.284 mg/m³ over a piece of plywood covering a floor flume where a machining center had been removed. Some of the peaks shown may have nothing to do with MWF. For example, in the "office conference room" there were particulate levels measured which could have been dust. There are other instances of exposure causing events noted in Figure 6.

The quartz crystal microbalance (QCM) cascade impactor results are shown in Figure 7 and Figure 8.

The results of the eight-stage (Anderson) impactor studies are shown in Figure 9 and Figure 10. For the Anderson, the LOD=0.1 mg/filter and the LOQ=0.3 mg/filter. The impactor located "between machining centers" yielded an overall concentration of 0.11 mg/m³ with 30 percent of the material deposited on the first filter or stage indicating that there was a heavy concentration of large particles. Without the material on the first filter, the concentration was 0.078 mg/m³. The corresponding numbers for near the "central cleaning" (Hydro) unit are 0.14 mg/m³, 37 percent and 0.087 mg/m³. Aerosols larger than 3-4 µm were present indicating exposures are probably due to uncontrolled operations.

DISCUSSION AND CONCLUSIONS

Phase 2 of the study indicated relatively low concentrations (less than 0.5 mg/m³) of total particulate and TEA. With both substances, the highest concentrations were found near the Hydromation (central cleaning) unit. This unit was apparently causing significant emissions of metalworking fluids into the plant's air. In order to reduce MWF mist concentrations throughout the plant this emission source was controlled.

Figure 5: Particulate concentrations measured with a hand-held aerosol monitor (HAM) worn by two operators in L-Shop who were also video monitored while performing their jobs, August 1996.





Figure 6: Total particulate concentrations measured with a Grimm portable dust monitor (PDM) in various locations in the plant and office areas and outside the plant, August 1996.





Figure 7: Total particulate near the "central cleaning (Hydro) unit''(C) collected with a quartz crystal microbahce cascade impactor, August 1996.





Figure 8: Total particulate "between machining centers"(B) collected by a quartz crystal microbalance cascade impactor, August 1996.





Figure 9: Total particulate "between machining centers''(B) collected by Anderson Impactor, August 1996. A total of 13.81 mg of material was collected.





Figure 10: Total particulate near the "central cleaning (Hydro) unit''(C) collected by Anderson Impactor, August 1996. A total of 16.87 mg of material was collected.

Phase 3 of the study demonstrated the effectiveness of installing air cleaners on the machining centers and improving mist control on the Hydromation unit. TEA concentrations were reduced four-to-ten fold. The total particulate concentrations were reduced 16 fold near the Hydromation unit, the location of highest concentration in the 'before' or Phase 2 study, and were significantly improved in the other area samples and in the personal samples.

Some major sources remain; they are the older machining centers which are not so well enclosed. These conditions should improve as older centers are replaced with new, more fully enclosed machining centers. It remains important to continue enclosure of the flumes and as much of the Hydromation unit operation as possible. The exposures to individual operators doing specific tasks were not large (less than 1.0 mg/m³) but could be decreased with improved ventilation and changes in work procedures.

The impactor results showed that there are still large particles being emitted into the work environment. The differences between the QCM and eight-stage studies, may indicate that the large particles are being emitted during the 1 hour air cleaner down time periods.

The installation of the air cleaners has resulted in a significant reduction in the TEA and total particulate concentrations in the plant work environment.

REFERENCES

1. Heitbrink WA, Spencer AB, Deye GJ [1996]. In-Depth Survey Report: Characterization of Metalworking Mists During the Evaluation of a Commercial Air Cleaner at Sauer Sundstrand, Ames, Iowa, June 8-14 and August 1-3, 1995. U.S. DHHS, PHS, CDC, NIOSH, NTIS Publication No. PB-96-191960.

2. Lapides MA [1994]. Cutting Fluids Expose Metal Workers to the Risk of Occupational Dermatitis. Occupational Health and Safety 63(4):82-86.

3. Kennedy SM, Greaves IA, Kriebel D, Eisen EA, Smith TJ, Woskie SR [1989]. Acute Pulmonary Responses Among Automobile Workers Exposed to Aerosols of Machining Fluids. Am J Ind Med 15:627-641.

4. Hendy MS, Beattie BE, Burge PS [1985]. Occupational Asthma Due to an Emulsified Oil Mist. British J Ind Med 42:51-54.

5. Eisen, EA [1995]. Extended Abstract: Case-Control Studies of Five Digestive Cancers and Exposure to Metalworking Fluids. Metalworking Fluids Symposium Final Programs and Abstracts, The Industrial Metalworking Environment: Assessment and Control. Dearborn, Michigan, pp. 19.

6. ACGIH [1996]. 1996 Threshold Limit Values for Chemical Substances and Physical Agents and Biological Exposure Indices. Cincinnati, OH: American Conference of Governmental Industrial Hygienists.

7. CFR. Code of Federal Regulations. Washington, DC: U.S. Government Printing Office, OffIce of the Federal Register. Labor, Table Z-1-A Limits for Air Contaminants: [29 CFR 1910.1000 (1992)].

8. ASHRAE [1996]. 1996 ASHRAE Handbook: Heating, Ventilating, and Air-Conditioning Systems and Equipment. AtIanta Georgia: American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc, pp. 24.4-24.5.

9. Fairchild CI, Wheat LD [1984]. Calibration and Evaluation of a Real-time Cascade Impactor. Am Ind Hyg. Assoc. J. 45(4):205-211.

10. NIOSH [1994]. NIOSH Manual of Analytical Methods, 4th Edition. Cincinnati, OH: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, Publication No. 94-113.

11. NIOSH [1977]. National Institute for Occupational Safety and Health: Occupational Exposure Sampling Strategy Manual. Cincinnati, OH: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control, National Institute for Occupational Safety and Health, Publication No. 77-173.

12. SAS/STAT User's Guide, Release 6.03 Edition, SAS Institute, Inc., SAS Circle, Box 8000, Cary, NC 27512-8000, 1988, pp. 570.

APPENDIX I