EVALUATION OF LOW- AND NON-VOC TECHNOLOGIES: APPLICATION TO SOUTH COAST AIR QUALITY MANAGEMENT DISTRICT CLEANING RULES

Prepared for:
The South Coast Air Quality Management District
Under SCAQMD Contract 98026
"Develop Low VOC Solvents"

Prepared by:
Katy Wolf and Mike Morris
Institute for Research and Technical Assistance

September, 1999

DISCLAIMER

This report was prepared as a result of work sponsored and paid for in whole by the South Coast Air Quality Management AQMD (AQMD). The opinions, findings, conclusions, and recommendations are those of the author and do not necessarily represent the views of AQMD. AQMD, its officers, employees, contractors, and subcontractors make no warranty, expressed or implied, and assume no legal liability for the information in this report. AQMD has not approved or disapproved this report, nor has AQMD passed upon the accuracy or adequacy of the information contained herein.

This analysis benefited considerably from the efforts of many persons within and outside the Institute for Research and Technical Assistance (IRTA). We would particularly like to acknowledge the valuable contributions made by the South Coast Air Quality Management District (SCAQMD) staff. We would also like to acknowledge the support we received from the three SCAQMD project monitors, Fred Minassian, Robert Pease and Abid Latif. We are especially grateful to the companies that agreed to test alternatives or use their case studies for this project. Finally, we appreciate the efforts of Irene Mah y Busch of IRTA in helping to prepare the document.

I. Introduction and Background ……………………………………………………………………………………. 1
1.1 New Case Study Companies…………………………………………………………………………………... 2
1.2 Existing Case Study Facilities ………………………………………………………………………………….. 2
1.3 Cleaning Agent Performance …………………………………………………………………………………... 4
1.4 Report Organization …………………………………………………………………………………………… 4
II. Rule 1171--Solvent Cleaning Operations ……………………………………………………………………….. 5
2.1 Surface Preparation Cleaning with Handwipe Method …………………………………………………………. 6
2.1.1 Charles Caine ………………………………………………………………………………………………..6
2.2 Product Manufacture--Electronic Components or Medical Devices ……………………………………………. 8
2.2.1 MiniMed and MRG ……………………………………………………………………………………...….. 8
2.2.1.1 External Product Manufacture ……………………………………………………………………………...9
2.2.1.2 Implantable Pumps ………………………………………………………………………………………..10
2.2.1.3 External Infusion Systems ………………………………………………………………………………... 11
2.2.2 Aerojet Electronic Systems ………………………………………………………………………………... 12
2.3 Repair and Maintenance Cleaning--General ………………………………………………………………….. 12
2.3.1 Paul's Transmission ………………………………………………………………………………………... 13
2.3.2 Newhall Carburetor ……………………………………………………………………………………….. 16
2.3.3 Rocketdyne Division of Boeing ……………………………………………………………………………..18
2.4 Repair and Maintenance Cleaning--Electrical Apparatus Components ……………………………………….. 20
2.4.1 Rocketdyne ……………………………………………………………………………………………….. 20
2.5 Cleaning of Coatings or Adhesives Application Equipment …………………..……………………………….. 22
2.5.1 Steelcase--Coating Application Equipment …………………………………………………………..…….. 23
2.5.2 Steelcase--Adhesive Application Equipment …………………………………………………………….…..26
2.5.3 Charles Caine--Coating Application Equipment …………………………………………………………….. 27
2.5.4 Hydro-Aire--Coating Application Equipment ………………………………………………………………..27
2.6 Cleaning of Ink Application Equipment--Lithographic and Letter Press ……………………………………….. 29
2.6.1 San Bernardino Sun ……………………………………………………………………………………….. 29
2.6.2 Nelson Name Plate …………………………………………………………………….………………….. 31
2.6.3 R R Donnelley & Sons …………………………………………………………………………………….. 33
2.7 Cleaning of Ink Application Equipment--Screen Printing ……………………………………….…………….. 36
2.7.1 Nelson Name Plate ……………………………………………………………………………………….. 36
III. Rule 1122--Solvent Degreasers ……………………………………………………………….…………….. 40
3.1 Batch Loaded Cold Cleaners and Conveyorized Cold Cleaners ……………………………………………… 40
3.1.1 S&H Machine, Inc. …………………………………………………….………………………………….. 41
3.1.2 Hydro-Aire ………………………………………………………………………………………………... 46
3.1.3 Litton Guidance & Control Systems ……………………………………………………………………….. 48
3.2 Vapor Degreasing ……………………………………………………….………………………………….. 52
3.2.1 General Metal Vapor Degreasing ………………………………………………………………………….. 54
3.2.1.1 California Electroplating Inc. …………………………………………………………………………….. 54
3.2.1.2 Hydro-Aire--Vapor Degreaser Conversions …………………………………………………………….. 56
3.2.1.3 Nelson Name Plate ………………………………………………………………………….………….. 58
3.2.1.4 Barranca Diamond Products …………………………………………………………………………….. 60
3.2.2 Electronics Cleaning ……………………………………………………………………………………….. 63
3.2.2.1 Hydro-Aire …………………………………………………….……………………………………….. 64
3.2.2.2 Aerojet Electronic Systems …………………………………………………………………………….... 66
3.2.3 Precision Cleaning ……………………………………………………………………..………………….. 67
3.2.3.1 Astro Pak ……………………………………………………………………………………………….. 67
3.2.3.2 Kaiser Electroprecision ………………………………………………………………………………….. 69
IV. Rule 1124--Aerospace Assembly and Component Manufacturing Operations ………………………………… 70
4.1 Hydro-Aire …………………………………………………………………………….…………………….. 70
4.2 Rocketdyne Division of Boeing ……………………………………………………………………………….. 71
4.2.1 Handwipe Cleaning ………………………………………………………..……………………………….. 71
4.2.2 Cleanliness Verification …………………………………………………………………………………….. 72
V. Rule 442 Usage of Solvents …………………………………………………….…………………………….. 74
5.1 Nelson Name Plate ………………………………………………………………………………………….. 74
5.2 Other Rule 442 Facilities …………………………………………………………………………………….. 75
VI. Other VOC Emissions ……………………………………………………………………………………….. 76
6.1 Cleanroom Maintenance ……………………………………….…………………………………………….. 76
6.2 Janitorial/Institutional Cleaning ……………………………………………………………………………….. 78
6.3 Machine Oil Dilution …………………………………………………..…………………………………….. 79
6.4 Vanishing Oil ……………………………………………………………….……………………………….. 79
6.5 Minimum VOC Limits ………………………………………………………….……………………………..79
VII. Results and Conclusions …………………………………………………………………………………….. 81
7.1 Results and Conclusions for Rule 1171 Cleaning Operations ………………………………………………….. 81
7.1.1 Surface Preparation ………………………………………………………….…………………………….. 81
7.1.2 Product Manufacture--Medical Devices …………………………………………...……………………….. 81
7.1.3 Prouct Manufacture--Electronic Components ………………………………………………………………. 81
7.1.4 Repair and Maintenance Cleaning--General ……………………………………………………..………….. 82
7.1.5 Repair and Maintenance Cleaning--Electrical Apparatus Components ………………………………………. 84
7.1.6 Cleaning of Coating or Adhesive Application Equipment ……………………………………………………. 84
7.1.7 Cleaning of Ink Application Equipment ……………………………………………………….…………….. 84
7.2 Results and Conclusions for Rule 1122 and NESHAP Cleaning Operations …………………………………… 84
7.2.1 Batch Loaded Cold Cleaners (BLCCs) …………………………………………………………………….. 84
7.2.2 Electronics Cleaning ………………………………………………………………….…………………….. 85
7.2.3 General Metal Vapor Degreasing ……………………………………………………………………….….. 85
7.2.4 Precision Cleaning ……………………………………………………………………….…………………. 85
7.2.5 Conclusions …………………………………………………………………………………………….….. 86
7.3 Results and Conclusions for Rule 1124 Cleaning Operations ……………………………………………….….. 86
7.4 Results and Conclusions for Rule 442 Operations ………………………………………………………….….. 86
7.5 Results and Conclusions for Other Cleaning Operations …………………………………………………….….. 86
7.6 Recommendations ………………………………………………………………………………………….…..86
VIII. Bibliography …………………………………………………………………….……………………….….. 92

Appendix A Products Cerftified as Clean Air Solvents by the South Coast Air Quality Management District

Appendix B Stand Alone Case Studies for Selected Companies
Transmission Shop Converts to Water-Based Cleaning Process
Carburetor Rebuilder Moves to Water-Based Cleaning
Small Burbank Machine Shop Adopts Water-Based Cleaning Process
Litton Converts Away from VOC Solvents
Small Plating Company Reduces Cost by Converting to Water Cleaning Process
Aerospace Company Converts to Water-Based Cleaning
Southern California Firm Converts to Water-Based Cleaning Process
Small Gardena Company Converts to Water Cleaning Process
Burbank Aerospace Firm Converts to Water-Based System
Precision Cleaning Company Converts from VOC Solvent to
Water-Based Cleaning System

Appendix C Material Safety Data Sheets for Rocketdyne Field Electrical
Handwipe Cleaning Tests

Appendix D Cleanroom Maintenance Article

FIGURES

Figure 2-1 Ultrasonic Cleaning Unit at Paul's Transmission ………………………………………….…………..14
Figure 2-2 Parts Cleaner at Paul's Transmission ……………………………………………………………….. 15
Figure 2-3 Ultrasonic Cleaning Unit at Newhall Carburetor ………………………………………..……….….. 17
Figure 2-4 Lithographic Press at Nelson Name Plate ……………………………………………………….….. 31
Figure 2-5 Screen Printing Operation at Nelson Name Plate ……….……………………………………….….. 37
Figure 2-6 Screen Recycling at Nelson Name Plate ………..……………………………………………….…..38
Figure 3-1 Large Parts Cleaner at S&H Machine ………………..………………………………………….…..42
Figure 3-2 Small Parts Cleaner at S&H Machine ……..…………………………………………………….…..43
Figure 3-3 Ultrasonic System at S&H Machine …………………………………………………………….….. 44
Figure 3-4 Parts Cleaner at Hydro-Aire …………………………………………………………………….…..46
Figure 3-5 Laser Gyro Showing Optical Components at Litton…………...………………………………….…..49
Figure 3-6 Frame Manufacturing Operation at Litton ……………………………………………………….…..50
Figure 3-7 Prism Operation at Litton ……………………………………………………………………….…..51
Figure 3-8 Ultrasonic System at California Electroplating …………………………..……………………….…..54
Figure 3-9 Ultrasonic System at Hydro-Aire ……………………………………………………………….…..56
Figure 3-10 Conveyorized Cleaning System at Nelson Name Plate ……………………………..………….…..59
Figure 3-11 Spray Cabinet at Barranca Diamond Products ……………………………………………..….…..62
Figure 3-12 Spray System at Hydro-Aire ………………………………………………………………….…..65

Table 1-1 New Case Study Companies and Cleaning Operations …………………………………..……….…..2
Table 1-2 Existing Case Study Companies and Cleaning Operations ………..……………………………….…..3
Table 2-1 Rule 1171 Cleaning Categories Selected for Study ……………………………………………….…..5
Table 2-2 Annual Cost Comparison for Charles Caine Handwipe Cleaning ………………………………….…..8
Table 2-3 Annual Cost Comparison for MiniMed Handwipe Applications ….……………………………….…..9
Table 2-4 Annual Cost Comparison for MiniMed Benchtop Cleaning ……………..……………………….…..10
Table 2-5 Annual Cost Comparison for MiniMed LCD Cleaning ……………………………………….….…..10
Table 2-6 Annual Cost Comparison for MiniMed Implantable Pumps ………………..…………………….…..11
Table 2-7 Annual Cost Comparison for MiniMed Cleanroom Maintenance …………..…………………….…..11
Table 2-8 Annual Cost Comparison for Paul's Transmission …………………………………….………….…..16
Table 2-9 Annual Cost Comparison for Newhall Carburetor ……………………………………………….…..18
Table 2-10 Annual Cost Comparison for Rocketdyne Field Handwipe Cleaning …………...……………….…..20
Table 2-11 Steelcase Application Equipment Cleaning Tests ……………………………………………….…..24
Table 2-12 Annual Cost Comparison for Steelcase Coating Spray Gun Cleaning ..………………………….…..25
Table 2-13 Annual Cost Comparison for Steelcase Adhesive Spray Gun Cleaning ………………………….…..26
Table 2-14 Annual Cost Comparison for Charles Caine Equipment Application Cleaning …………………...…..27
Table 2-15 Hydro-Aire Application Equipment Cleaning Tests ……………………………………………..…..28
Table 2-16 Annual Cost Comparison for Hydro-Aire Application Equipment Cleaning ………………..…….…..29
Table 2-17 Annual Cost Comparison for San Bernardino Sun Lithographic Printing Cleaning …….………….…..30
Table 2-18 Annual Cost Comparison for Nelson Name Plate Lithographic Printing Cleaning …………….….…..32
Table 2-19 Annual Cost Comparison for R R Donnelley & Sons Lithographic Printing Cleaning …………….…..36
Table 2-20 Annual Cost Comparison for Nelson Name Plate Screen Printing Cleaning …….……………….…..38
Table 3-1 Annual Cost Comparison for S&H Machine Batch Loaded Cold Cleaning ……………………….…..45
Table 3-2 Annual Cost Comparison for Hydro-Aire Batch Loaded Cold Cleaning ………………………….…..47
Table 3-3 Annual Cost Comparison for Litton Batch Loaded Cold Cleaning .……………………………….…..52
Table 3-4 Annual Cost Comparison for California Electroplating Vapor Degreaser Conversion ….………….…..56
Table 3-5 Annual Cost Comparison for Hydro-Aire Vapor Degreaser Conversion …………………...…….…..58
Table 3-6 Annual Cost Comparison for Nelson Name Plate Vapor Degreaser Conversion …...…………….…..60
Table 3-7 Annual Cost Comparison for Barranca Diamond Products Vapor Degreaser Conversion ………...…..63
Table 3-8 Annual Cost Comparison for Hydro-Aire Printed Circuit Board Leaning …...…………………….…..66
Table 3-9 Annual Cost Comparison for Astro Pak Vapor Degreaser Conversion …………………..……….…..68
Table 5-1 Annual Cost Comparison for Nelson Name Plate Photoresist Stripping Conversion ……...……….…..75
Table 7-1 Investigated Cleaning Categories ……………………..………………………………………….…..82
Table 7-2 Cleaning Rule Exemptions ……………………………………………………………………….…..83
Table 7-3 Project Recommendations ……………………………………………………………………….…..91

I. INTRODUCTION AND BACKGROUND

Volatile Organic Compound (VOC) emissions from solvent cleaning and degreasing operations contribute significantly to the South Coast Air Basin's emission inventory. The South Coast Air Quality Management District (SCAQMD or District) adopted an Air Quality Management Plan (AQMP) in 1997. This AQMP calls for significant reductions in VOC emissions from cleaning and degreasing operations by 2010 to achieve attainment status.

Some of the District's rules that focus on cleaning and degreasing had future compliance limits for which technology had not yet been specified. In order to help develop low- or non-VOC technologies to comply with these provisions and to satisfy the AQMP's goals, the District contracted with the Institute for Research and Technical Assistance (IRTA). Under the contract, IRTA investigated and tested low- and non-VOC alternatives in a variety of cleaning and degreasing processes. The aim was to identify technologies that could be substituted for high VOC technologies used today in many types of cleaning.

At the beginning of the two-year project, IRTA and the District staff identified the cleaning applications where more work and development and demonstration of low VOC technologies was needed. Once this screening was accomplished, nine companies in the South Coast Basin were selected to participate in the project. The cleaning operations in these nine facilities were evaluated and low- or non-VOC alternatives were tested. In several cases, the firms involved in the testing converted their processes to the alternatives that were found to be the most promising during the testing. In these instances, the companies implemented the alternatives and the technical feasibility and cost of the alternatives could be determined directly. In other cases, where the facility did not end up adopting the alternative, the technical feasibility and cost of the alternatives was assessed through analysis using the best available data.

The project was designed so that promising areas of future VOC reductions could be identified. In addition, the District and IRTA wanted to provide guidance to industry on the technologies that could be used to comply with what were future rule effective dates at the project's initiation. Promising areas of VOC reductions were considered to be categories with either a high VOC emission inventory or a high VOC content limit. The inventory values for the cleaning categories were derived using SCAQMD numbers or SCAQMD reports where possible. In some cases, new categories were defined and the derivation of the inventory numbers is explained in the text.

IRTA has worked with many companies in the Basin and has assisted them in converting to low- or non-VOC alternatives with funding from other sources. In addition to the nine companies participating in the project, IRTA and the District selected ten companies that IRTA had converted or was in the process of converting to cleaning alternatives. Two additional companies that had made conversions several years ago were also selected. As part of the project, IRTA evaluated the technical feasibility and cost of the conversions and included them in this report.

Part of the project involved examining the cleaning agents that are available for satisfying low VOC limits. The District has established a Clean Air Solvent Certification Program and several of the cleaning agents that have received certification were tested in the course of the project. Cleaning agents can be certified by the District if they meet several criteria. Some of the criteria are that the cleaning agent contain 50 grams per liter VOC content or less as used, that the chemical not be classified as ozone depleting or global warming compounds or as Hazardous Air Pollutants (HAPs). The District has certified more than 100 Clean Air Solvents and more than 75 percent of them contain 25 grams per liter VOC or less. Appendix A lists several of the products that have been certified as Clean Air Solvents. The program is ongoing and new products are added to the list as they are qualified.

1.1. New Case Study Companies

The nine companies that agreed to participate in the project are listed in Table 1-1. The cleaning operations that were the focus of the project are also listed in the table and the applicable SCAQMD rule is identified. These rules are discussed in more detail in subsequent sections of this document. IRTA's work with the nine companies involved investigation, testing and analysis of 13 cleaning operations.

1.2. Existing Case Study Facilities

Case studies for 12 companies that had already been completed or were being completed under other IRTA projects were also included in the SCAQMD project. In a few cases, IRTA did not work with the facility but performed technical feasibility and cost analysis of their conversions. In other cases, IRTA assisted the company in converting to an alternative process. The companies are shown in Table 1-2, together with the cleaning operations that were the subject of the technical assistance work and analysis. Sixteen cleaning operations at these companies were analyzed.

1.3. Cleaning Agent Performance

Performance of the cleaning agent at each facility in each application was judged on a case-by-case basis. In each case, the plant personnel provided information on their requirements for the cleaning process. In nearly all cases, the major criterion was if the cleaned part could successfully be put through the next process step. For cleaning prior to coating or plating, for instance, the criteria would be that the part have sufficient adhesion for the coating or be able to accept plating. In terms of performance, a cleaning agent was judged as successful if it cleaned the part as well as or better than the cleaning agent the company used in the process currently.

Plant personnel also had other criteria that related to safety and regulations. They did not want to use cleaning agents that were toxic and posed a risk to the workers or that appeared on various toxics lists. In order to minimize the risks of the cleaning agents to the workers and the surrounding community, a hierarchy was used for the testing. If water-based cleaners could be used in the process, then water-based cleaners without solvent additives were tested first. If these did not work effectively, water-based cleaners with solvent additives were tested. In some cases, blends of water-based cleaners or plain deionized water with solvents like isopropyl alcohol or acetone were tested. Isopropyl alcohol and acetone were selected because they are relatively low in toxicity and acetone is exempt from VOC regulations. If the application could not tolerate water, acetone was tested. In a few cases, other chemicals exempt from VOC regulations with relatively low toxicity were tested.

1.4. Report Organization

The report is organized into chapters that focus on a specific District rule. Section II addresses the cleaning operations that are covered by Rule 1171 "Solvent Cleaning Operations". Section III provides details on the operations that are regulated under Rule 1122 "Solvent Degreasers" or the Halogenated Solvent Cleaning NESHAP. Section IV highlights the operations that are covered under Rule 1124 "Aerospace Assembly and Component Manufacturing Operations." Section V addresses the operations covered by Rule 442 "Usage of Solvents." Section VI focuses on the operations that are not covered presently in District rules. Section VII summarizes the results of the project and makes recommendations for cleaning categories that could be targeted for regulation in the future. Finally, a bibliography is provided in Section VIII.

Appendix A is a list of the products certified as Clean Air Solvents by the District. Appendix B includes several short stand-alone case studies for some of the companies that participated in the project. They highlight one or more conversions the company has made and are useful as examples for similar facilities considering conversions. Appendix C presents Material Safety Data Sheets (MSDSs) for several aerosol formulations tested for energized electrical equipment at one of the companies that participated in the project. Appendix D includes a paper on alternatives for cleanroom maintenance that was published in 1997.

II. RULE 1171--SOLVENT CLEANING OPERATIONS

SCAQMD Rule 1171 regulates solvents and cleaning devices/methods that are used in production, repair and maintenance or servicing of parts, products, tools, machinery, equipment or general work areas. The rule sets VOC limits for a variety of different categories of cleaning solvents and provides a list of approved cleaning devices and methods.

In the course of this project, IRTA and SCAQMD staff selected several categories of cleaning in the rule where the VOC limits were high. IRTA identified facilities that wanted to participate in the testing of low- or non-VOC alternatives. SCAQMD staff and IRTA also agreed to include information on conversions for some categories of Rule 1171 where lower VOC limits had already been adopted. The areas of focus in this project for Rule 1171 are listed in Table 2-1 together with the current VOC and vapor pressure limits currently specified in Rule 1171. The table also provides estimates for the emission inventory for each of the categories.

On June 13, 1997, the District amended Rule 1171 to require that solvents used in repair and maintenance cleaning must have a VOC content of 50 grams per liter or less by January 1, 1999. Since this rule change affected thousands of small businesses, IRTA and SCAQMD staff decided to also focus on a few operations in this category where it was anticipated that conversion would be more challenging. IRTA was working with several facilities to assist them in converting to water-based cleaners for repair and maintenance cleaning and two of these companies were included in the analysis presented here.

The balance of this section discusses each of the different categories in Table 2-1. It also presents details about the testing at each of the facilities and the conversions some of them made to low- and non-VOC technologies.

2.1. Surface Preparation Cleaning With Handwipe Method

Surface preparation is the removal of contaminants such as dust, oil, soil, grease etc. prior to coating, adhesive or ink application. Surface preparation is most commonly carried out using the handwipe method.

Handwipe or wipe cleaning refers to cleaning operations that are done by hand often with solvent laden rags or Q-tips. These operations are labor intensive and they contribute substantially to VOC emissions. Numerous facilities in the Basin perform handwipe cleaning.

The VOC limit in this category has been set at 70 grams per liter for several years. It was not an issue in the past because most companies relied on the exempt chemicals 1,1,1-trichloroethane (TCA) and trichloroetrifluoroethane (CFC-113) to perform surface preparation handwipe cleaning operations. A production ban on TCA and CFC-113 went into effect in January 1, 1996. A congressional tax added to these ozone depleting solvents also raised their prices substantially. Over the last few years, many firms have converted away from the two solvents. Some of these companies converted to high VOC solvents and are not aware that they are not in compliance with the rule.

IRTA worked with one company a few years ago to assist the firm in converting away from TCA in surface preparation handwipe operations. The information on the conversion for this company is presented here.

2.1.1. Charles Caine

Charles Caine is a jobshop that provides powder and liquid coating services to a variety of aerospace and non-aerospace customers. In these coating operations, the parts must be clean for good paint adhesion. Charles Caine, like most companies, cleans their parts prior to applying the coatings. The company used a TCA vapor degreaser for cleaning the smaller parts prior to coating. The larger parts that required coating, like lamp posts for example, were handwiped with TCA because they were too large to fit in the degreaser. The Halogenated Solvent Degreasing NESHAP became effective on existing operations in December, 1997 and the price of TCA was very high. Charles Caine would have had to upgrade the degreaser to comply with the NESHAP requirements and it was at this stage that the company began evaluating other options.

Charles Caine investigated water-based bath cleaning, water-based handwipe cleaning and solvent handwipe cleaning. The company decided against water-based bath cleaning because a capital investment in new equipment would have been necessary. At that stage, the company decided to do all their cleaning by hand. The first alternative that was tested was lacquer thinner which performed fairly well and the company converted to it for a short time.

IRTA began working with the company to test a blend of acetone and water for surface preparation handwipe cleaning. The acetone/water blend performed as well as the lacquer thinner and the facility converted to the blend for 80 percent of their production. IRTA also assisted the company in testing a water-based neutral cleaner, Daraclean 236, which is made by W.R. Grace. This cleaner also performed well and the company converted to it in a 4 percent blend with water for cleaning 20 percent of their production. The Daraclean 236 is certified by SCAQMD as a Clean Air Solvent. Charles Caine could not use the Daraclean 236 for all applications because on some of the parts with crevices, all of the oil could not be removed effectively with the cleaner.

If Charles Caine had upgraded their degreaser to meet the NESHAP standards, the capital cost would have been $10,000. The upgrade would have involved increasing the freeboard and adding a refrigerated freeboard chiller. Assuming a 10 year lifetime for the degreaser and a 6 percent cost of capital, the annual cost of the upgrade is $1,360.

In 1996, the price of TCA was about $27.28 per gallon. Charles Caine purchased 715 gallons of TCA annually. The total annual cost for purchasing TCA is $19,505.

Charles Caine uses 10 gallons per week or 520 gallons per year of handwipe solvent. About 80 percent of the solvent or 416 gallons per year is a 10 percent acetone/90 percent water blend and 20 percent or 104 gallons per year is the Daraclean 236. The cost of acetone is $3.31 per gallon so the cost of the acetone/water blend is 33 cents per gallon. The cost of the Daraclean 236 is $10 per gallon so the cost of the 24:1 blend with water is 40 cents per gallon. The total annual cost of the two water-based cleaners is $179.

Charles Caine estimates that the use of the lacquer thinner also amounts to 520 gallons per year. The total annual cost of purchasing lacquer thinner, at a price of $3.10 per gallon, is $1,612.

Charles Caine employed one full time person to operate the vapor degreaser. When the degreaser was shut down, this worker and half of an additional worker were assigned to handwipe cleaning. Charles Caine's labor rate is $14 per hour. Assuming the personnel work 260 days per year, the labor cost before the degreaser was shut down was $29,120; the labor cost for the two employees after the shutdown is $43,680.

The shutdown of the degreaser eliminated the electricity cost for running the unit. The degreaser had an 18 kW heater and a 1 HP chiller. The degreaser was left on 24 hours per day. Assuming the heater and chiller cycled on about one-fourth of the time and that Charles Caine's electrical rate is 12 cents per kWh, the total electricity cost for operating the degreaser was $4,928.

Charles Caine canceled the SCAQMD permit when the degreaser was shut down. The permit renewal fee is $179. The company emitted about 536 gallons of TCA per year. They avoided a TCA emissions fee of 3 cents per pound or $77 annually through the conversion. There would be no emission fees for the lacquer thinner emissions because Charles Caine is below the 4 ton per year threshold.

There is no hazardous waste disposal cost for the degreaser or the handwipe operations. The recycler paid a credit for the waste TCA and charged a fee for the still bottom disposal. There was no net cost for the disposal.

Table 2-2 summarizes and compares the costs for using the vapor degreaser, using the lacquer thinner in a handwipe operation and using the water-based handwipe solvents.

The cost of using the vapor degreaser is higher than the cost of handwiping the parts. Converting to the current process--use of the acetone/water and Daraclean 236--reduced the total annual cost of cleaning by 21 percent. The labor cost increased but this cost was offset by the high cost of TCA.

The cost of using lacquer thinner and the water-based handwipe cleaners are comparable. Converting to the water-based process reduced lacquer thinner VOC emissions by 520 gallons or about 1.7 tons annually.

Charles Caine is typical of many companies that used TCA for handwipe cleaning. Often they converted to VOC solvents. This case demonstrates that alternatives to VOC solvents are available and perform well.

2.2. Product Manufacture--Electronic Components or Medical Devices

IRTA began working with one medical device manufacturer, MiniMed, during this project to test alternatives. By the end of the project, some of the work was completed with MiniMed and some with Mann Research Group (MRG), a MiniMed spin-off company. MiniMed and MRG used a variety of VOC and other solvents in small quantities in a number of applications. IRTA also worked with Aerojet Electronic Systems to analyze the costs of their conversion five years ago away from CFC-113 and TCA to a water-based process for printed circuit board assembly. This process conversion is discussed in Section III under the vapor degreasing--electronics category. Aerojet also converted to a water-based system for rework of some of their boards and this conversion is discussed briefly here.

The total emission inventory for this category in the Basin, as shown in Table 2-1, is 1.5 tons per day. Electronic Components account for 0.8 tons per day and medical devices account for about 0.7 tons per day.

2.2.1. MiniMed and MRG

Medical device manufacturers include companies that make medical instruments, implantable devices, surgical equipment, dental and medical supplies and pharmaceuticals. MiniMed is a medical device manufacturer that makes external devices. The company uses small amounts of solvents in a variety of different applications. MiniMed is much like other medical device manufacturers which have a range of operations, each using a small amount of solvent. Even though these companies generally have a large number of employees producing their high value goods, emissions from medical device manufacturers are generally low. These manufacturers are regulated by the Food and Drug Administration (FDA) and process changes are difficult to implement because of a lengthy approval process; approval may require as long as five years.

IRTA, MiniMed and MRG focused on three operations during this project. First, MiniMed uses VOC solvents in the manufacture of "external" products, products that are not implanted. Second, MRG uses CFC-113 in small quantities in implantable pumps. Third, MiniMed uses VOC solvents for external infusion systems.

2.2.1.1. External Product Manufacture

In external product manufacture, MiniMed uses two VOC solvents. IPA is used the most extensively. MiniMed uses 132 gallons per year for handwiping the interior and exterior of parts and for benchtop cleaning. About five gallons per year of heptane is used to clean the LCD displays.

IRTA and MiniMed structured a program to test alternatives to IPA in the handwiping application. One of the alternatives that was investigated is acetone. Again, the acetone was effective for the cleaning operation. Again, however, the workers did not like the odor.

IRTA suggested that MiniMed test several different blends of deionized (D.I.) water with IPA. The results of this testing indicated that a 50 percent D.I./50 percent IPA blend worked as well as pure IPA for parts cleaning. IRTA also suggested that MiniMed try a water-based cleaner. The cleaner was tested but left an unacceptable residue.

IRTA also suggested that MiniMed test blends of IPA and water for benchtop cleaning. IRTA had performed a study a few years ago that is discussed in more detail in Section V. That study found that 90 percent D.I. water/10 percent IPA and 100 percent D.I. performed as well as higher percentage IPA blends for particulate control. MiniMed tested the IPA blends and found that 90 percent D.I./10 percent IPA performed effectively.

In the cost analysis, IRTA assumed that half the IPA used in external product manufacture is used for handwiping parts and half is used for benchtop cleaning. This indicates that 66 gallons per year are used in each application.

The cost of IPA for MiniMed is assumed to be $9 per gallon. The alcohol used by medical device manufacturers is generally sterile and its price is much higher than IPA sold for other industrial applications. MiniMed generates D.I. for many operations on-site and it was assumed that the D.I. used for the alternatives had no incremental cost. The cost of the 90 percent D.I./10 percent IPA is 90 cents per gallon. The cost of acetone is placed at $8 per gallon.

For purposes of analysis, it is assumed that the use of acetone and IPA is equal. It is also assumed that there would be no change in labor or utilities from the substitutions.

Table 2-3 compares the annual cost of using pure IPA, 50 percent D.I./50 percent IPA and acetone for handwiping the parts assuming that MiniMed uses 66 gallons of IPA annually for that purpose. Through adoption of the 50 percent D.I./50 percent IPA blend, MiniMed could reduce VOC emissions by 218 pounds or 0.11 tons per year. Adoption of acetone would reduce VOC emissions by 436 pounds or 0.22 tons per year.

Table 2-4 compares the cost of pure IPA and the 90 percent D.I./10 percent IPA blend for benchtop cleaning assuming that MiniMed uses 66 gallons of IPA annually in that operation. Use of the D.I./IPA blend would reduce VOC emissions by 392 pounds or 0.20 tons per year.

IRTA suggested that MiniMed test acetone as a replacement for heptane in the LCD display application. MiniMed did try acetone for this application and it performed adequately. The workers, however, did not like the smell of acetone so the company continued to use heptane.

The cost of heptane is higher than the cost of acetone. IRTA estimates the cost of sterile heptane at about $15 per gallon. Table 2-5 presents the cost comparison for heptane and acetone for cleaning LCD displays.

The cost of using acetone is lower than the cost of using heptane. VOC emissions would be reduced by about 29 pounds per year if the acetone alternative were implemented.

2.2.1.2 Implantable Pumps

MRG manufactures implantable pumps for delivery of insulin. This application is not a cleaning process. Even so, IRTA decided it would be worthwhile to investigate alternatives because many medical device manufacturers use small amounts of solvents in a variety of products for a range of purposes not classified as cleaning.

CFC-113 is used in small amounts in the devices to provide the proper pressure to cause a bellows to release the insulin when it is needed. The chemical used for this purpose must have the required vapor pressure at 37 degrees C, the body temperature. CFC-113's vapor pressure is -4 psi at 37 degrees C. According to MRG, the ideal vapor pressure for an alternative would be -6 psi or 437 mm Hg.

Although MRG uses very small amounts of CFC-113 and has some in reserve, eventually an alternative will be needed for this operation. CFC-113 production was banned in the U.S. in 1996 and, even though there is an inventory, the supply will eventually be exhausted. Moreover, the devices containing CFC-113 must be labeled to alert buyers that they contain ozone depleting substances and this is a public relations problem. Thus it is important to MRG to identify potential alternatives.

IRTA identified several potential alternatives to CFC-113 that had the proper physical properties. Most of the alternatives were VOCs and most had flashpoints. Several had toxicity problems. The best alternative identified by IRTA was a hydrofluorocarbon, HFC-43-10. Its physical properties appeared better than those of CFC-113 for this application. The vapor pressure of HFC-43-10 at body temperature is -6 psi, the ideal pressure specified for the chemical used in this application. IRTA suggested that MRG test the alternative and a test stand is being built for that purpose. Even if the alternative is suitable, however, MRG would require FDA approval to use it.

IRTA performed a cost analysis for the alternative. About 10 cubic centimeters of CFC-113 is used in each pump and about 500 pumps are manufactured each year. The total use of CFC-13 is 5 liters or about 1.32 gallons annually. The current price of CFC-113 is $200 per gallon. The price of HFC-43-10 is $225 per gallon. The total cost of purchasing CFC-113 for the operation is $264 annually; the cost of purchasing HFC-43-10 is $297 annually. Table 2-6 shows the cost comparison.

CFC-113 and HFC-43-10 are both exempt from VOC regulations. HFC-43-10 appears to be a viable alternative in this application based on its properties. The emission reduction of CFC-113, an ozone depleter, through the substitution is very small. Nevertheless, the application is a very important one. From a VOC standpoint, use of the HFC-43-10 would prevent MRG from converting to another chemical that is a VOC.

2.2.1.3. External Infusion Systems

The major use of VOC solvents in this application is for cleanroom maintenance. MiniMed estimates that about 10 gallons per month are used for this purpose. This application is analyzed here even though cleanroom maintenance is not currently classified under Rule 1171. It may fall under the category of repair and maintenance cleaning or it may fall under the category of janitorial cleaning. The application is discussed in more detail in Section VI later. MiniMed currently uses pure IPA for cleanroom maintenance even though biocidal control is not necessary in these particular applications. As discussed in Section VI, IRTA performed a cleanroom maintenance study which found that plain D.I. water was more effective than pure IPA for controlling particulates. The medical industry has determined that, where biocidal control is required, a 70 percent IPA/30 percent D.I. blend is adequate.

The cost analysis is performed for the case where biocidal control is required and biocidal control is not required. The IPA price is assumed to be $9 per gallon. It is assumed that the incremental cost of generating D.I. water is for cleanroom maintenance is negligible.

Table 2-7 shows the cost comparison for MiniMed's use of pure IPA, use of the 70 percent IPA/30 percent D.I. blend and use of plain D.I. water. MiniMed's VOC emissions are reduced by 238 pounds or 0.1 tons per year by substituting the 70 percent IPA/30 percent D.I. blend; they are reduced by 792 pounds or about 0.4 tons per year through the conversion to plain D.I. water.

MiniMed is representative of many medical device manufacturers. These companies generally use a fair amount of IPA for sterility and biocidal purposes for cleanroom maintenance, handwipe cleaning and cold cleaning. They use small amounts of other VOC solvents for other handwipe applications. Except for the cleanroom maintenance application, alternatives must be identified on a case-by-case basis.

This case study demonstrates that acetone may be suitable for some medical device cleaning operations where biocidal control is not necessary. Some of the less critical applications could also use a 70 percent/30 percent IPA blend with D.I. water. Finally, in some cases, where biocidal control is not necessary, plain D.I. water may be a suitable IPA substitute.

2.2.2. Aerojet Electronic Systems

In the printed circuit board assembly process, components are soldered to boards. Flux is applied to the boards before soldering to effect heat transfer and to make the solder flow more easily. The components are soldered to the board and the flux and any other contaminants that are present are cleaned. Like most other companies that assemble printed circuit boards, Aerojet used CFC-113 and TCA often in blends with alcohol in a vapor degreasing process for cleaning the boards. At times when the boards are assembled, they need to be reworked for a variety of reasons. Companies generally used CFC-113 or TCA combined with small amounts of alcohol in the rework process. The rework was generally performed at a benchtop with the solvent in small squeeze bottles.

About five years ago, Aerojet converted from the vapor degreasers to a water-based process for cleaning the flux from the boards. The conversion is described in more detail in Section III later. The machine Aerojet purchased is a conveyorized design with a wash, rinse and dry section. For rework of the boards at benchtop, Aerojet, like nearly all other manufacturers, converted to isopropyl alcohol (IPA). For some of the rework, however, Aerojet is able to clean the boards in the new machine which uses a water-based cleaner.

Aerojet cannot clean all the reworked boards in the conveyorized water-based cleaning machine because of logistics. Some of the rework activities are performed in other buildings or rooms in the same building that are far from the cleaning system. It would not be practical to travel a long distance to clean the reworked boards. Aerojet has just purchased two new batch printed circuit board cleaning units. The company may be able to use these in different locations and clean a larger percent of their reworked boards with the water-based cleaner.

2.3. Repair and Maintenance Cleaning--General

The June 13, 1997 amendments to Rule 1171 specified that solvents used for repair and maintenance cleaning must have a VOC content of 50 grams per liter or less by January 1, 1999. This provision affected 40,000 parts cleaners in the South Coast Basin. About 25,000 of the parts cleaners were in auto repair facilities and 15,000 were in industrial facilities. The reduction in VOC emissions for repair and maintenance cleaning amounted to about 16 tons per day.

In 1997 and 1998, IRTA performed a project with funding from EPA, Southern California Edison and the Pollution Prevention Center to assist auto repair and industrial facilities in converting to water-based cleaners for repair and maintenance cleaning. As part of this work, IRTA assisted a transmission and a carburetor repair shop in testing water-based cleaners and adopting systems that were effective for their cleaning needs. Because these applications involved complex parts, they were thought to pose the most challenge for the conversion in general repair and maintenance cleaning. IRTA and SCAQMD staff decided to include the analysis of the conversion for these two facilities here.

In addition to the analysis for these two auto repair shops, analysis for an aerospace company, the Rocketdyne Division of Boeing, is included here. Rocketdyne tested alternatives for field repair and maintenance handwipe cleaning.

2.3.1. Paul's Transmission

Paul's Transmission actually consist of two facilities, owned and operated by the Griffitts family, that repair and rebuild transmissions. One facility is in Santa Monica and the other is located in Culver City. Both of these operations used the same cleaning systems with mineral spirits and both converted to the same water-based cleaning systems. The analysis below was performed for only one of the facilities, the shop in Santa Monica.

In 1998, Paul's Transmission was using two mineral spirits sink-on-a-drum units for cleaning all kinds of parts including the valve bodies that the shop repairs. The cost of these two mineral spirits units, including servicing and disposal, was $115 per month or $1,380 per year. The shop also had a large spray cabinet that relied on a water-based cleaner. The wastewater from the spray cabinet is discharged to a clarifier. The shop technicians spent about four hours per day cleaning parts in the two sink-on-a-drum units. Assuming a five-day week, the total labor cost for cleaning with the solvents was $15,600 annually.

To comply with the SCAQMD Rule 1171 requirements, the shop purchased an ultrasonic cleaning unit for $4,500 for cleaning the valve bodies and other parts. This price also included the first charge of water-based cleaning formulation, Daraclean 257, which is certified as a Clean Air Solvent. Figure 2-1 shows a picture of the ultrasonic system. Assuming a five-year equipment lifetime and that Mr. Griffitts paid cash, the annual cost of the unit amounts to $900. The unit is changed out every two months at a cost of cleaner of $55. Small amounts of cleaner are added over the period to replace evaporation. Assuming this constitutes about 10 percent of the total bath, the replacement cost is $5.50 every two months. The total cost of cleaning agent for this unit is $363 per year.

Ultrasonic water-based cleaning is more effective than mineral spirits for penetrating the passages of the valve bodies and other parts. Furthermore, because the unit is automated, the labor cost for cleaning is reduced substantially. Mr. Griffitts estimates the cleaning time of the ultrasonic unit at 2 hours per day. Since the unit is automated, however, the technician spends only about 10 percent of his time, or 12 minutes per day, in cleaning. Assuming a labor rate of $15 per hour, the labor cost for operating the ultrasonic unit is $780 per year.

The shop replaced one of the mineral spirits sink-on-a-drum units with a water-based sink-on-a-drum unit at a cost of $800. This parts cleaner is shown in Figure 2-2. The company also recently purchased a filtration system for $250 to extend the bath life of the water-based cleaner. The total cost of this system was $1,050. Assuming a five-year equipment lifetime, the annual cost of the system is $210.

The system uses a water-based cleaner, Daraclean 236, which is certified as a Clean Air Solvent. The price of the cleaning concentrate is $12.90 per gallon. The capacity of the system is 30 gallons and about 30 percent of the cleaner is required for effective cleaning. Each time the bath is changed out, the cost of the cleaner is $116. Before the filter was added, the bath lasted only a month; Mr. Griffitts estimates a three-month bath life since the filter has been added. The cost of changing out the bath four times a year amounts to $464.

Assuming that about 10 percent of the cleaner is required for make-up for evaporation, the total annual cost of the cleaner is $510.

The technicians spent two hours per day cleaning in the mineral spirits unit. Water-based cleaning with a sink-on-a-drum requires somewhat more labor, perhaps 20 percent more. The time spent in cleaning with the water-based cleaning unit is 2.4 hours per day. At a labor cost of $15 per hour, this amounts to $9,360.

The shop kept one of the mineral spirits units for cleaning valve bodies that have attached solenoids and electronic switches. Both types of parts are electronic components and Rule 1171 provides a higher VOC limit, 900 grams per liter and a vapor pressure of 20 mm Hg or less, for electrical apparatus components. Because this unit is used much less than it was previously, the annual cost, including servicing and disposal, is about $345 per year. The technicians spend an estimated 30 minutes per day cleaning parts in this unit. The total labor cost for this activity, assuming a labor rate of $15 per hour, is $1,950 per year.

Figure 2-2 Parts Cleaner at Paul's Transmission

 In earlier studies, IRTA estimated the electrical cost of using a mineral spirits sink-on-a-drum at $5 per month or $60 per year. For the two mineral spirits parts cleaners, the electrical cost was $120 annually. The electrical cost paid by the shop increased by about $25 per month or $300 per year after installation of the two new water-based cleaning systems. This is the electrical cost for continuing to use one of the mineral spirits parts cleaners, the new water-based sink-on-a-drum unit and the ultrasonic unit. The total electrical cost under the water cleaning scenario is $420.

The shop does not have to pay for disposal of the water-based cleaners. When the water-based cleaner is spent, it is poured into the large spray cabinet where it is used to replace evaporation. As mentioned above, the spray cabinet charge goes to the clarifier.

Table 2-8 summarizes the cost comparison for mineral spirits and water-based cleaning for Paul's Transmission. The total cost of using the mineral spirits was about $17,000. The cost of using the water-based cleaners, at about $15,000, is 13 percent lower. The cost of purchasing the new water-based cleaning units is more than offset by the savings in labor from using the ultrasonic cleaning unit.

Appendix B includes a stand-alone case study for Paul's Transmission.

2.3.2. Newhall Carburetor

Carburetors mix fuel vapor and air before combustion in gasoline engines. In the early 1980s, fuel injectors replaced most carburetors. Many older cars and classic cars, however, still utilize carburetors. Nearly all carburetors are rebuilt and/or repaired. The dirty carburetors are contaminated with fuel deposits, oil, grease, road grime, and carbon deposits. Carburetors have narrow valves and blind passages which makes cleaning difficult.

Traditionally, carburetors were cleaned prior to repair or rebuilding in an agitated tank with a mixture of mineral spirits and various types of chlorinated solvents. The carburetor was removed from the vehicle and placed directly into an unheated agitated tank containing the solvent mixture. The tank was agitated with an electrical motor that rotated the carburetor 180 degrees back and forth. The carburetors were left in the agitated tank for several hours. They were removed and touch up hand cleaning with aerosol spray solvents, mineral spirits and wipe rags completed the cleaning process.

IRTA began working with Newhall Carburetor to identify a suitable alternative for this cleaning process. The shop typically works on several carburetors per day. They also do general auto repair but specialize in carburetor repair for automobiles, trucks and boats.

Two sink-on-a-drum systems were investigated first. One of these was an Zymo enzyme system that the shop had purchased earlier to replace a mineral spirits parts cleaner. The second system was a parts cleaner with a more aggressive cleaning agent. Neither of these systems proved adequate for cleaning the carburetors.

IRTA arranged for an equipment supplier to build a heated agitated tank similar to the traditional solvent carburetor cleaning unit. The cleaner was alkaline and fairly aggressive. The temperature of the bath was set at 160 degrees F. The carburetor was left in the agitated bath overnight. This time the results were better but the carburetors were not as clean as they were when cleaned in the solvent carburetor cleaner.

Finally, an ultrasonic cleaning system was tested. The ultrasonic cleaner was manufactured by Alpha Cleaning Systems. The tank was heated to 150 degrees F and an aggressive alkaline cleaner was used for the testing. The carburetors were cleaned for about thirty minutes. Since the metal carburetors retain heat well and the bath is heated to a high temperature, the carburetors flash dried. This process was capable of cleaning the carburetors as well as or better than the solvent agitation system with subsequent hand cleaning.

Newhall Carburetor decided to purchase an ultrasonic cleaning system for cleaning the carburetors. The system tested in the facility was large but Alpha Cleaning Systems also offered a smaller unit that was 18 inches cubed. This is the unit the facility decided to purchase. A picture of the unit is shown in Figure 2-3.

Newhall Carburetor leased the solvent carburetor cleaning unit for $75 per month or $900 per year. This also included the cost of chemical and the cost of disposal. The firm converted to an ultrasonic cleaning unit at a cost of $3,000 or $600 per year assuming a five-year equipment lifetime. The cleaner the shop is using, Daraclean 257, is certified as a Clean Air Solvent. The 25-gallon bath is changed out six times a year and 10 percent additional cleaner is required as makeup. The total annual water-based cleaner cost, assuming a price of $15 per gallon, is $248.

 With the carburetor cleaning unit, the shop spent 20 minutes cleaning each carburetor by hand and they typically clean 10 carburetors per week. At a labor rate of $22 per hour, the labor cost with mineral spirits was $3,667 annually. With the water-based cleaner, since the system is automated, the labor cost is one-tenth the cost or $367 annually.

The electrical cost for the mineral spirits unit was estimated at $5 per month or $60 per year. The electrical cost of the ultrasonic unit is assumed to be $50 per month or $600 per year.

The disposal cost for the water-based cleaner, assuming six changeouts per year and a disposal cost of $200 per drum for three drums, is $600.

The total annual cost of using the water-based cleaner is $1,815. This is nearly one-third the $4,627 annual cost of using mineral spirits.

Table 2-9 shows the cost comparison for mineral spirits and the water-based cleaner. The cost of cleaning with the water-based cleaner is about half the cost of cleaning with the mineral spirits. The major cost savings come from a reduced labor cost because the ultrasonic system is automated.

A stand-alone case study for Newhall Carburetor is presented in Appendix B.

2.3.3. Rocketdyne Division of Boeing

When Rule 1171 was being amended to require solvents with a VOC content of 50 grams per liter in repair and maintenance cleaning, the industry brought field handwipe cleaning operations to the District's attention. The District agreed that field handwipe cleaning was more challenging and allowed an exemption until January 1, 2001. For purposes of the exemption, field handwipe cleaning operations means "cleaning operations conducted in areas located at least 1000 feet outside the operator's facility boundary, for the repair of mobile equipment, non-road equipment, and other internal combustion engine driven equipment including, but not limited to, buses, trucks, tractors, forklifts, pumps, and heavy-duty construction equipment."

Rocketdyne performs field handwipe cleaning operations on many types of equipment. Their cleaning includes handwiping of stationary devices like compressors that are too large or impractical to disconnect or disassemble. Some companies use mineral spirits or other VOC solvents for their field handwipe operations. In the past, Rocketdyne used TCA for field handwipe cleaning.

A few years ago, Rocketdyne converted to a water-based cleaner for field handwipe cleaning. The company uses a product called Mean Green, a full strength concentrate alkaline cleaner that is diluted with water. Rocketdyne mixes the diluted Mean Green into spray bottles and directly applies it to areas to be cleaned. The cleaner is used on air conditioners, vehicles, benches and tables, virtually any equipment that needs to be cleaned without being moved with the exception of electrical equipment.

During the testing with Rocketdyne, IRTA suggested the workers try a water-based aerosol cleaner to compare it with the Mean Green. IRTA provided them with the only water-based aerosol on the market at the time. The Rocketdyne workers found the performance of the aerosol cleaner, Mirachem Parts Washer in a Can, to be effective. They also liked the aerosol delivery system better that Mean Green spray bottles because the Mirachem cleaner did not need to be diluted and it was pressurized.

In Rule 1171, aerosol products are not subject to the 50 gram per liter VOC cutoff level for repair and maintenance cleaning if 160 fluid ounces or less are used per day per facility. The Mirachem aerosol product contains a glycol ether an the concentrate, which is used in the aerosol, has a VOC content of 80 grams per liter. The company also uses a VOC propellant which is 3 percent by weight. Although companies are permitted to use 160 ounces of aerosols per day with a VOC content higher than 50 grams per liter, IRTA does not recommend that they do so. A few other water-based aerosol products are now on the market. They are propelled by carbon dioxide and have a lower VOC content than the cleaner tested here.

Both the Mean Green and the Mirachem water-based cleaners perform well for field maintenance. There is a trade-off between ease-of-use and cost, but shop managers can weigh the merits of each. What the testing demonstrates is that water-based cleaners are suitable for field handwipe. The Mean Green must be diluted and used in the less convenient spray bottle delivery system and the aerosol can packaging is more expensive. There is no real difference in labor expended or quantity used to complete a cleaning task. The only real differences are in cost and air emissions.

The price of the Mean Green concentrate is $5.21 per gallon. About half the Mean Green used by Rocketdyne is diluted in a 50 percent blend with water; about half is diluted in a 25 percent blend with water. Diluted to 50 percent Mean Green/50 percent water, the cost is $2.60 per gallon; diluted to 25 percent Mean Green/75 percent water, the cost is $1.30 per gallon. Rocketdyne uses about three gallons of the Mean Green concentrate each month for field maintenance. The total number of gallons of diluted Mean Green used is 7 gallons per month or 84 gallons per year. The total annual cost of using the Mean Green amounts to $188.

The price of the Mirachem is $3.33 per can or $23.68 per gallon. Assuming that 84 gallons of the Mirachem aerosol are also used each year, the total cost of using this alternative is $1,989.

Some companies use mineral spirits for field handwipe operations. This analysis compares mineral spirits cleaning with water-based cleaning. Assuming that 84 gallons of mineral spirits were used for the field maintenance and assuming the cost is $2 per gallon, the total annual cost for using mineral spirits would amount to $168.

The costs of the three options are summarized in the Table 2-10.

The values show that the cost of using mineral spirits is lower than the cost of using either of the water-based cleaners. The cost of using the Mean Green, however, is only about 12 percent higher than the cost of using mineral spirits. The cost of using the Mirachem aerosol product is more than 10 times higher than the cost of using the Mean Green. Users definitely pay more for the convenience of using an aerosol product.

In terms of emissions, the Mirachem aerosol product is about 100 grams per liter VOC. The Mean Green is an alkaline based product with no VOC content according to the MSDS. The mineral spirits VOC content is 780 grams per liter. For a company like Rocketdyne using mineral spirits, a reduction in VOC emissions of 546 pounds or 0.27 tons per year could be achieved through adoption of a water-based cleaner.

2.4. Repair and Maintenance Cleaning--Electrical Apparatus Components

Solvents that are used for repair and maintenance cleaning of electrical apparatus components must have a VOC content of 900 grams per liter or less and a vapor pressure of 20 mm Hg or less. Most electrical devices cannot be cleaned with solvents with a flash point, particularly if they are energized. For many years, TCA and CFC-113 were used by virtually all companies for cleaning electrical apparatus components. These two ozone depleting solvents were effective and they had no flash point.

In 1996, production of CFC-113 and TCA was banned worldwide because the chemicals contribute to ozone depletion. A congressional tax placed on the chemicals increased their cost substantially. Users of non-energized electrical equipment, like generators, for example, substituted VOC solvents like terpenes and mineral spirits for the CFC-113 and TCA. Most users with energized electrical equipment substituted a new alternative, HCFC-141b, for the TCA and the CFC-113. This HCFC has an ozone depletion potential that is roughly the same as that of TCA. For that reason, EPA banned its use in cleaning applications except handwipe operations. All uses of HCFC-141b will be banned in 2003.

At this stage, most companies use HCFC-141b, generally in aerosol form, for cleaning energized field electrical equipment. The solvent is fairly aggressive and, like CFC-113 and TCA, it has no flash point. Once the chemical is banned in 2003, these companies will have to find an alternative. As part of this project, IRTA decided to investigate this type of cleaning and test alternatives to determine the best path once HCFC-141b is banned. The Rocketdyne Division of Boeing agreed to work with IRTA to test potential alternatives.

2.4.1. Rocketdyne

Like other companies, Rocketdyne used TCA and CFC-113 aerosol cleaners for cleaning live field electrical equipment. Like other companies, Rocketdyne also converted to the HCFC-141b aerosol cleaners some years ago. The types of components cleaned by Rocketdyne with the aerosol includes transformers, very large diffusion pumps, motor bearings and housings, electrical parts of welding guns, energized circuits and high voltage equipment. Some of these components can be de-energized, taken out of service, dismantled and cleaned with water-based cleaners. The others that cannot be dismantled were cleaned with the aerosol cleaners in the field.

In this application, the challenge is to identify effective solvents that have no flash point and have low or no VOC content. The requirement that the solvents have no flash point indicates that a halogenated solvent must be used. These halogenated solvents can be combined with non-halogenated solvents as long as the VOC content is not high and as long as the blend does not have a flash point.

With these restrictions, IRTA identified six products for testing at Rocketdyne. IRTA also asked DuPont to formulate one additional cleaner that appeared promising. The seven products contain hydrofluoroethers (HFE 7100 and 7200), a hydrofluorocarbon (HFC-43-10) or a hydrochlorofluorcarbon (HCFC-225). HCFC-225 contributes minimally to ozone depletion and it is scheduled to be banned over the next 20 years. The HFEs and HFC-43-10 do not contribute to ozone depletion since they contain no chlorine; both do contribute somewhat to global warming, however. All of the chemicals have no flash point and all are exempt from VOC regulations.

The HFEs and HFC-43-10 are very nonaggressive cleaners and they are generally combined with other solvents to boost their cleaning capability. HCFC-225 is somewhat more aggressive but it is still only a mild cleaning agent. Again, blends of HCFC-225 with more aggressive cleaners are generally used.

One solvent that is often combined with these cleaners, particularly the HFEs and HFC-43-10, is 1,2-trans dichloroethylene (DCE). The latter chemical has a flash point so it cannot be used alone. When it is combined in a 50 to 60 percent DCE blend with the HFE or HFC, the mixture does not have a flash point and it is a fairly aggressive cleaner. IRTA did not test these blends for two reasons. First, DCE is a VOC and the mixtures of DCE with the HFEs and HFC have high VOC content. Second, although DCE has never been tested for chronic toxicity, it has a structure that indicates it could be a carcinogen. Rocketdyne personnel would not allow blends containing DCE to be tested for this reason.

One other halogenated solvent that does not have a flash point is available. n-Propyl bromide (NPB) is a chemical that is relatively new to the market. It contributes slightly to ozone depletion but its main drawback is that it is a VOC. Because NPB is a VOC, because it can contain a contaminant, 2-bromopropane, that is extremely toxic and because it may be toxic itself, IRTA did not test NPB in this application. Again, Rocketdyne personnel were not willing to test the chemical because of toxicity concerns.

IRTA identified seven cleaners that might be suitable for the field electrical handwipe applications at Rocketdyne. The material safety data sheets (MSDSs) for these products are included in Appendix C. They are:

o a DuPont product called Vertrel KCD-9550 containing HFC-43-10 and acetone. This is the product IRTA asked DuPont to specially formulate for the testing.
o a Miller-Stephenson product called MS-760 Vertrel X-P10 which contains HFC-43-10 and isopropyl alcohol (IPA). The propellant is HFC-134a which is exempt from VOC regulations.
o a Miller-Stephenson product called MS-730 which contains the two HFEs, IPA and HFC-134a as the propellant.
o a Tech Spray product called HFE Cleaner Degreaser which contains the two HFEs, IPA, acetone and HFC-134a as the propellant.
o a Tech Spray product called Asahiklin 225 1663 aerosol which contains two isomers of HCFC-225 and HFC-134a and carbon dioxide as propellants.
o a Tech Spray product called Blue Shower II which contains the two isomers of HCFC-225, methanol, ethanol, nitromethane as a stabilizer and HFC-134a as the propellant.
o a Tech Spray product called Asahiklin 225 1668 aerosol which contains the two isomers of HCFC-225, acetone and HFC-134a and carbon dioxide as propellants.

IRTA presented the products to Rocketdyne and the Rocketdyne personnel further screened them for testing. Rocketdyne wanted to eliminate the last three listed products from consideration. HCFC-225 is relatively toxic and the worker exposure level of the chemical is set at 50 ppm. Rocketdyne personnel believed that workers would be exposed to higher levels of the chemical if it was applied in aerosol form. They decided to test only the first four products.

IRTA arranged for samples of the four cleaners to be sent to Rocketdyne. Over the next month or so, the Rocketdyne workers responsible for field maintenance tested the aerosol products for cleaning different types of energized equipment. The results were disappointing. The cleaners were simply not aggressive enough to adequately clean the parts.

About a year ago, Rocketdyne substantially reduced the use of the HCFC-141b because it contributed to ozone depletion. Rocketdyne reasoned that the chemical would be completely banned from use in 2003 and decided to reduce its use earlier. After Rocketdyne reduced their use of HCFC-141b, the field maintenance staff stopped cleaning some of the energized electrical equipment. Testing of the four alternative cleaners showed that none was at all effective for the cleaning task. Rocketdyne has decided that the best option for the company is to not clean some of their energized electrical equipment and to test the HCFC-225 alternative in the future.

The implications of not cleaning the equipment are not known so a cost analysis cannot be performed. As mentioned earlier, some of the equipment can be de-energized, dismantled and brought to the facility for cleaning with water-based cleaners periodically. Some other large pieces of equipment, like generators for example, can be de-energized and cleaned with water-based cleaners in the field.

Companies other than Rocketdyne may be willing to ignore the toxicity problems with DCE, NPB and HCFC-225 which are more aggressive cleaners. The other problem with DCE and NPB, however, is that they are VOCs and their use would increase the emissions of VOCs for facilities that are currently using HCFC-141b. HCFC-225 is not a VOC but it contributes to ozone depletion. It is not scheduled to be banned for at least 15 years and overall it may be the best alternative.

2.5. Cleaning of Coatings or Adhesives Application Equipment

Nearly all liquid spray coating and adhesive operations require application equipment cleaning. Spray equipment is commonly cleaned after most color or product changes and at the end of a shift. Cleaning consists of removing paint or adhesive from the nozzle orifices, flushing the supply lines and removing excess paint from the interior and exterior of the spray guns.

Spray guns are cleaned by wiping the exterior with solvent and then immersing the spray gun in solvent or flushing the inside passages by spraying the solvent through them. SCAQMD regulations require flushing to be done in a closed container. Companies commonly use an enclosed spray gun cleaner or spray into a closed waste drum to meet this requirement. Many shops use open buckets to soak and rinse their equipment and some shops spray the cleaning solvent into the spray booth. Facilities which use open buckets or spray cleaning solvent into the booth tend to use much more clean-up solvent than a shop which recirculates the cleaning solvent in an enclosed container. In a typical spray station, about one quart of cleaning solvent is used each time the gun is cleaned.

Adhesive gun cleaning generally has lower emissions than coating gun cleaning. Normally the guns are cleaned at the end of the day or sometimes at end of the week. IRTA has worked extensively in adhesives and has visited companies all over the country that apply adhesives with spray equipment. Most shops, whether they use waterborne or solventborne adhesives, use very little cleaning solvent. Some types of waterborne adhesives can simply be peeled off. The tips or nozzles of the guns used to apply adhesives are generally the only parts that require cleaning every day. Guns used with solventborne adhesives are periodically disassembled and cleaned. If a large amount of adhesive sets-up in the spray equipment or lines, however, it must be taken out of service and cleanup may require a full day. Many waterborne adhesive users use water for cleanup. The solventborne adhesive users generally used the solvent that is the carrier in their adhesive for cleanup.

IRTA tested alternatives cleaning agents for coating equipment clean-up at three facilities: Steelcase in Tustin; Charles Caine in Los Angeles; and Hydro-Aire in Burbank. Steelcase is a large furniture manufacturer that applies solvent-borne coatings to metal substrates. Charles Caine is a job shop that applies high-solids and water-borne coatings to metal. Hydro-Aire applies aerospace coatings, primarily primers and topcoats, to metal aircraft parts. Steelcase uses electrostatic spray equipment while Charles Caine and Hydro-Aire use HVLP spray guns.

The testing protocol used the company's current gun cleaning solvent as the baseline for the performance evaluation. IRTA, with the company's workers, cleaned the exterior and interior (cups) of the guns and flushed the lines. IRTA and the workers evaluated the cleaning time and effectiveness relative to the baseline cleaning agent. The details and results of the test are summarized for each facility below.

During the project, IRTA also gathered cost information on an adhesive application equipment cleanup comparison from Steelcase. That comparison is also summarized below.

2.5.1. Steelcase--Coating Application Equipment

At Steelcase, the cleaning tests were conducted for two types of coatings that are used by the facility extensively. Both coatings are baking enamels supplied by Sherwin Williams. Steelcase currently uses xylene to clean the coating application equipment. Four different alternative cleaning agents were tested. These included an alkaline water-based cleaner made by W.R. Grace called Daraclean 257 which is certified as a Clean Air Solvent, a blend of acetone with this water-based cleaner, pure acetone and a blend of acetone with xylene. The coatings, cleaning agents and the results of the tests are summarized in Table 2-11.

For both types of coatings, the best performing cleaning of all five cleaning agents, including xylene, the baseline cleaner, was acetone. The blend of acetone and xylene performed better than plain xylene. The blend of acetone with the Daraclean 257 performed as well as xylene. The Daraclean 257 with water foamed during the cleaning. The cost analysis was performed for xylene as the baseline cleaner, for the 100 percent acetone alternative and for the 50 percent acetone/50 percent Daraclean 257 alternative.

Steelcase purchases the xylene used currently for application equipment cleanup for $2.00 per gallon. The company currently uses 94 gallons of xylene per day. Assuming 260 days per year of operation, this amounts to 24,440 gallons per year. The total annual cost for purchasing the xylene is $48,880.

Two alternatives were analyzed. The first alternative is to substitute 100 percent acetone for the xylene currently used by Steelcase. The price of acetone is estimated at $2.10 per gallon. Steelcase would likely use about 10 percent more acetone than xylene because of acetone's higher vapor pressure. Under this assumption, the annual use of acetone would amount to 26,884 gallons. The total annual cost of purchasing acetone is $56,456.

The second alternative that was analyzed is to substitute the 50 percent acetone/50 percent Daraclean 257 for xylene. The price of Daraclean 257 is about $12.90 per gallon. Although the blend of acetone and Daraclean 257 contains acetone which is more volatile than xylene, it also contains the water-based cleaner concentrate which is substantially less volatile than xylene. IRTA has found in the past that the water retards the evaporation of the acetone. For purposes of analysis, it was assumed that the use of the blend would be 25 percent less than the use of xylene. The annual cost of purchasing 18,330 gallons of the blend each year, at $7.50 per gallon, is $134,475.

Steelcase handles the spent cleaner as hazardous waste. A company representative estimates that about 60 percent of the xylene that is purchased, or 14,664 gallons per year, ends up in the waste. It is also estimated that the waste contains about 15 percent solids. On this basis, the total amount of hazardous waste generated is 16,864 gallons per year. Steelcase pays $1.50 per gallon for waste disposal. The waste disposal costs for the xylene waste amount to $25,296. The cost for disposal of the acetone waste would be the same as for the xylene waste. Although acetone purchases are higher because more acetone is emitted, the same volume ends up in the hazardous waste. The waste generated from using the blend of acetone and Daraclean 257 would carry the same cost for disposal. Because the hazardous waste costs are the same for all three cleaners, they are not included in the summary table below.

Steelcase pays $292.80 per ton for annual VOC emissions to the AQMD. Emissions of xylene are 40 percent of purchases or about 35 tons per year. The annual cost of emission fees is $10,248. No District fees are levied on acetone because it is not classified as a VOC.

Table 2-12 summarizes the cost comparison for the three cleaning agents.

Table 2-12
Annual Cost Comparison for Steelcase Coating Spray Gun Cleaning

The cost analysis shows that the cost to Steelcase for using acetone is lower by about 5 percent than the cost of using xylene. A conversion to acetone would reduce annual VOC emissions by about 35 tons. The cost of using the acetone/Daraclean 257 blend is more than double the cost of using xylene.

IRTA has not evaluated the long term effects of acetone, water or aqueous cleaners on spray equipment. Some users claim that the water will corrode the equipment or interfere with the next coating that is applied. However, many shops already use water exclusively to clean waterborne coatings and any corrosive effects have not been an issue. It is also hard to imagine that a few drops (or maybe ounces) of water would ruin the paint the company applies next. Often, a painter will adjust gun settings and color and flow test on a test panel or back wall of the booth when he/she begins painting and any remaining water in the line would be sprayed out at that time.

Steelcase raised two concerns about using acetone in place of xylene. The first issue is whether polar solvents like acetone with very low flash points can be used with electrostatic application equipment. One spray gun manufacturer indicates that a modification package is available that modifies the spray gun for use with polar solvents in either coatings or cleanup solvents. The second issue is the flammability of acetone. Because of the flammability rating of the chemical, the amount that could be stored on-site is limited by the local fire department. Use of the low flash point solvent might also change Steelcase's insurance carrier provisions.

The acetone blend with Daraclean 257 worked well but the cost of using this blend would be very high. The advantage of this blend is reduced flammability and worker exposure and it might not have the problems with storage and insurance fees that pure acetone would have. IRTA did not test acetone blends with Daraclean 257 and plain water. Neither did IRTA test a 50 percent acetone/50 percent plain water blend. The lower Daraclean 257 blends and the plain water blend would still offer the advantages of reduced flammability and worker exposure but would have a much lower cost. Other water-based cleaners in more dilute form that are less expensive might also perform adequately.

Steelcase cannot use a water-based cleaner for cleanup of their xylene based coatings unless they purchase new application equipment. The company has an electrostatic spray gun that is not suitable for water. Companies can purchase electrostatic spray guns that are designed to spray waterborne coatings and these guns can be cleaned with water-based cleaners. Although water-based cleaners would not be suitable for Steelcase's operation, they might be appropriate for other companies with HVLP guns or electrostatic equipment designed for use with waterborne coatings.

2.5.2. Steelcase--Adhesive Application Equipment

The adhesives that Steelcase used in the past were based on TCA. When the chemical was banned and the price increased through a congressional tax, TCA based adhesives were no longer available. At that stage, Steelcase began testing alternative adhesive and has generally converted to water-based and hot melt products. When the company used TCA based adhesives, they cleaned the application equipment with TCA. The cost analysis in this section compares the cost of using TCA and water for cleanup of application equipment.

At the time Steelcase used TCA, the cost was $5.50 per gallon. About 20 gallons per month were used for cleaning the adhesive application equipment. The total annual cost of the TCA was $1,320. After the conversion to the alternative adhesives, plain water was used for the cleanup.

About 10 minutes per day were devoted to cleanup of the adhesives application equipment when Steelcase used TCA. Steelcase noted an increase of about 25 percent in the time spent in the cleanup activities with the conversion to water-based adhesives. The labor rate used was $14 per hour. The annual labor cost for TCA cleanup was $607. The annual labor cost after the conversion to water-based adhesives is $759.

At the time Steelcase used TCA, the company paid $1.50 per gallon for hazardous waste disposal. It is assumed that half the TCA was emitted and half remained as hazardous waste. It is also assumed that the hazardous waste volume is equal to the usage of TCA because the waste solids double the waste volume. On this basis, the annual hazardous waste disposal cost for TCA was $360. There is no hazardous waste generated from cleanup with water.

SCAQMD emission fees for TCA amounted to 3 cents per pound at the time the company used the chemical. Emissions of TCA from the cleanup operation are 10 gallons per month or 120 gallons per year. The annual emission fees for TCA were about $40.

Table 2-13 summarizes the cost comparison for the adhesive operation.

The table indicates that the cost of using water for cleanup of application equipment is much less than the cost of using TCA. Although TCA is not a VOC, emissions of the ozone depleting chemical were reduced by 1,320 pounds or 0.66 tons per year. Moreover, if Steelcase had not converted to waterborne adhesives, the company might have converted to a VOC solvent for cleanup.

2.5.3. Charles Caine--Coating Application Equipment

Charles Caine is a jobshop that provides a coating service to customers. The company has a powder coating operation and a liquid coating operation. In the liquid coating operation, Charles Caine uses an HVLP spray gun. Some of the coatings used by Charles Caine are waterborne and some are solventborne.

The company uses water for cleanup of the application equipment after spraying waterborne coatings. The company uses lacquer thinner for application equipment cleaning after spraying the solventborne coatings. The types of solventborne coatings Charles Caine applies include high solids and epoxy urethanes. The cleanup solvents were tested on a two-part polyurethane coating.

IRTA and Charles Caine tested acetone as an alternative to the lacquer thinner for application equipment cleaning. The acetone performed well. The company uses approximately 5 gallons per month of cleanup solvent and the workers noticed no difference in use with the acetone.

Charles Caine purchases lacquer thinner at a cost of $3.10 per gallon. Assuming the 5 gallon per month usage level, the annual cost for purchasing lacquer thinner is $186. Charles Caine pays $3.31 per gallon for purchasing acetone. If the facility converted to acetone, the annual cost for purchasing the chemical would amount to $199.

There is no change in the cost of waste disposal for lacquer thinner and acetone. Charles Caine is below the 4 ton per year cutoff level for paying VOC emission fees to the SCAQMD.

Table 2-14 shows the cost comparison for lacquer thinner and acetone.

The total annual cost of using acetone is seven percent higher than the cost of using lacquer thinner. Assuming that emissions of lacquer thinner are half the use, a conversion to acetone would reduce Charles Caine's VOC emissions by 198 pounds annually.

2.5.4. Hydro-Aire--Coating Application Equipment

Hydro-Aire is an aerospace subcontractor. Like other aerospace subcontractors, the company applies aerospace topcoats and primers. The application equipment used by the company is an HVLP spray gun. Hydro-Aire was using Aero MEK in an enclosed spray gun cleaner for cleaning their application equipment.

IRTA and the Hydro-Aire worker that applies the coatings tested several different alternative cleaners on two types of coatings, an epoxy primer and a polyurethane topcoat. Hydro-Aire uses several different types of coatings but the topcoat was judged to be representative and the primer that was selected for testing, an impact resistant coating, is very hard to clean. For this primer and this primer alone, Hydro-Aire uses the epoxy primer thinner for cleaning; the Aero MEK is ineffective for cleaning the primer. Table 2-15 below summarizes the different cleaning agents that were tested on the two types of coatings.

The tests indicated that acetone and the blend of acetone and Aero MEK were better cleaners than the plain Aero MEK. In fact, the Aero MEK was not effective and was not used for cleaning the primer. This suggests that pure acetone could be used for cleaning Hydro-Aire's coatings. Hydro-Aire agreed to test the 50% acetone/50% Aero MEK blend over the longer term in their enclosed spray gun cleaning unit. At the time of this writing, the tests were still underway.

Hydro-Aire purchases Aero MEK at a price of $3.40 per gallon and acetone at a price of $3.31 per gallon. Hydro-Aire estimates that the company uses about 10 gallons per month of Aero MEK. The company indicates that there has been no increase in use for the blend. It is also assumed that the use of acetone would not increase. The total annual cost for purchasing Aero MEK, the blend of Aero MEK and acetone and plain acetone is $408, $404 and $397 respectively.

The cost for disposal of all three types of wastes is the same and it is not included in the cost comparison. Hydro-Aire pays SCAQMD emission fees for VOCs of $292.80 per ton annually. Assuming that half the solvent used is emitted and half is in the hazardous waste, the annual emission fees for Aero MEK and the blend of Aero MEK and acetone are $58 and $29 respectively. There are no emission fees for acetone since the chemical is not classified as a VOC.

Table 2-16 shows the cost comparison for the three cleanup solvents.

The cost of using acetone for cleaning the application equipment is the lowest of the three options. The cost of using acetone is about 15 percent less than the cost of using Aero MEK. By converting from Aero MEK to acetone, Hydro-Aire could reduce their VOC emissions by about 0.2 tons annually.

2.6. Cleaning of Ink Application Equipment--Lithographic and Letter Press

Hundred, perhaps thousands, of printers operate in the South Coast Basin. There are several different types of printing operations including lithographic, flexographic, gravure and screen printing. The most widely used process and the process that uses and emits the most cleaning solvent is lithographic printing. Lithographic printing is synonymous with offset printing. Both sheetfed and web (or continuous) presses are common. Typical uses for sheetfed lithographic printing include books, catalogs, posters, boxes and labels. Typical uses for web lithographic presses are newspapers, books and magazines.

The rule regulates cleaning solvents that are used to clean ink application equipment. In this light, ink application equipment includes the blankets, rollers, guards, rails, pumps, and troughs.

During this project, IRTA worked with two facilities, RR Donnelley & Sons, a large lithographic printer, and Nelson Name Plate, a name plate manufacturer, to test alternatives for cleaning ink application equipment. IRTA also collected information from a third facility, the San Bernardino Sun, on the paper's conversion to a water-based cleaner. The testing and conversion at each of these companies is described in more detail below.

2.6.1. San Bernardino Sun

The San Bernardino Sun prints the San Bernardino Sun and USA Today newspapers. The company operates 18 4-color offset presses. They use soy-based inks for newspaper printing. On a daily basis, the company has several different printing jobs on each press. They typically print 250,000 48-page papers with 21.5 inches per page each day. The presses are cleaned after every printing job.

Mirachem, a company that formulates and sells water-based cleaners, had designed one of its cleaning agents for ink cleanup. Mirachem told IRTA that the San Bernardino Sun used their water cleaner which has been certified as a Clean Air Solvent. The company uses a parts washer with a water-based cleaner, Mirachem Pressroom Cleaner, at 20 percent concentration. The San Bernardino Sun had converted to the water-based cleaner as a replacement for mineral spirits. The parts cleaners are used for cleaning parts that are removed from the press which includes the ink pans and other items during repair of the press.

The company had also tried the Pressroom Cleaner for handwipe cleaning on the press. At any dilution, they found that the Pressroom Cleaner did not perform as well as the mineral spirits blend they used. At full strength, the water-based cleaner performed as well as or better than the mineral sprits. The company did not convert to the water-based cleaner, however, because at 100 percent concentration, it would have been much more expensive than the mineral spirits blend. The workers also tested a 20 percent concentration Pressroom Cleaner on rags for handwipe press cleaning. They indicated that there was an oil residue left behind and they had to do follow-up cleaning with the mineral spirits.

The company reduced their use of the mineral spirits blend by 60 gallons per week by using the water-based cleaner in the parts cleaner. The price of the mineral spirits blend is $2 per gallon. The annual cost for purchasing the solvent amounts to $6,240.

The Mirachem parts cleaner is leased as part of a servicing arrangement. The cost of the water-based cleaner is included in the servicing price. The cost of the service amounts to $180 per month or $2,160 per year.

The parts cleaner is heated so the company experienced an electricity use increase. IRTA estimates that the electricity cost of one parts cleaner amounts to about $10 per month or $120 per year.

The San Bernardino Sun recycles all of their ink and solvents so there are no disposal fees from using the solvent. The waste disposal of the water-based cleaner is included in the servicing fee for the parts cleaner.

The conversion to the water-based cleaner reduced VOC emissions by 60 gallons per week or 10.92 tons annually. The company paid a SCAQMD fee for emitting the VOC solvents. These fees are $292.80 per ton. The annual emission fee amounted to $3,197.

Table 2-17 compares the cost of using the mineral spirits and water-based cleaner for the off-press cleaning.

The conversion to the water-based parts cleaner reduced the San Bernardino Sun's costs substantially, by about 76 percent. VOC emissions were reduced by 10.92 tons annually through the conversion.

2.6.2. Nelson Name Plate

IRTA worked with Nelson to identify and implement a new process for cleanup of their lithographic press operation. This operation is performed as part of the manufacturing process for some of the name plates. The inks that Nelson uses are solventborne, air dry lithographic metal deco inks.

For many years, Nelson had been using TCA as a blanket wash, a roller wash and a plate cleaner in between runs on the lithographic press. A picture of this press is shown in Figure 2-4. Because of the production ban on TCA and because the price had increased substantially, Nelson wanted to identify, test and implement an alternative.

The cleaners most of the vendors were offering as alternatives had a VOC content of about 900 grams per liter. Nelson had a facilitywide limit on their VOC emissions and, in the conversion away from the exempt TCA, wanted to find a non- or low-VOC alternative. IRTA tested various water-based materials including water/acetone blends. IRTA also had Nelson's vendor formulate blends of acetone with the mineral spirit based cleaner that was being offered for press cleaning. In April, 1996, Nelson converted to a blend of acetone and mineral spirits that had a VOC content of 2.68 pounds per gallon or 326 grams per liter. Over the next year or so, additional testing allowed Nelson to convert to a higher content acetone blend with a VOC content of 1.60 pounds per gallon.

The use of TCA as a cleanup solvent amounted to about 20 gallons per month or 240 gallons per year for this operation. The price of TCA was $16.15 per gallon. This leads to an annual TCA cost of $3,876. The usage of solvent did not change with the conversion to the two VOC solvent blends and the cleanup labor remained the same. The cost of the 2.68 pound per gallon blend was $3.20 per gallon. The annual cost of the conversion was $768. The cost of the 1.60 pound per gallon blend was a little higher, at $3.70 per gallon. The annual cost of the operation with this solvent is $888.

Disposal fees have remained the same for this operation. When TCA was used, the rags used for the wipe cleaning were sent to an industrial laundry and returned to Nelson for reuse. The same number of rags were used after conversion to the VOC-based solvents and the disposal was handled in the same way.

There is some information that suggests that use of acetone on a press as a blanket wash may damage the blanket and that the press would require earlier replacement. This does not appear to be the case at Nelson. The company has been using an acetone-based blanket wash for more than two years and has not noticed any damage.

Nelson paid fees of three cents per pound for emissions of TCA to the SCAQMD. This totaled $79 annually. Emissions of VOCs are charged by SCAQMD at a rate of $292.80 per ton. Fees for emissions of the 2.68 pound per gallon blend amounted to about $94 per year. Fees for emissions of the 1.60 pound per gallon blend are lower, at $56 per year.

Table 2-18 compares the cost of the conversion from TCA to the higher and lower VOC content blanket and roller wash. The costs for using the VOC solvent blends are much lower than the cost of using TCA. The cost of using the 1.60 pound per gallon blend is 76 percent lower than the cost of using TCA.

If Nelson had converted from TCA to a 900 gram per liter or 7.5 pound per gallon compliant cleaner, their emissions from this operation would have been 1,800 pounds or 0.9 tons per year. By using the 2.68 pound per gallon VOC blend, Nelson's emissions from this operation amounted to only 643 pounds or 0.3 tons annually. In converting to the 1.60 pound per gallon VOC cleaner, Nelson reduced their VOC emissions to 384 pounds or 0.19 tons per year. The cost of the transition from the 2.68 pound per gallon cleaner to the 1.6 pound per gallon cleaner was $633 per ton. Nelson reduced their overall cost substantially by converting away from TCA to the 1.6 pound per gallon cleaner.

2.6.3. R R Donnelley & Sons

R R Donnelley & Sons is located in Torrance. The company is one of the largest lithographic printers in the country. They print newspaper inserts and magazines which is considered to be higher quality printing than newspaper printing. They have a control device for their ink emissions but do not consider emissions from cleanup activities to be controlled.

R R Donnelley workers do the bulk of their cleaning by hand. Parts are cleaned on a table with rags and squirt bottles of mineral spirits. When IRTA began working with the company, they also had seven mineral spirits parts cleaners located throughout the facility. After evaluating their cleaning operations, R R Donnelley decided that the parts cleaners were not necessary and they discontinued the service. At that stage, all of the off-press cleaning was performed by handwiping.

When IRTA began working with R R Donnelley, the company had just started experimenting with the Mirachem Pressroom Cleaner. IRTA first suggested that R R Donnelley try to use the Mirachem cleaner for the handwipe operations to replace mineral spirits. IRTA and the workers conducted side-by-side testing with different dilutions of Pressroom Cleaner versus mineral spirits. The tests were performed with concentrations of 10, 20, 30 and 50 percent of the Pressroom cleaner in water. The cleaners were compared on the basis of the time it took to clean the part and the amount of energy the worker had to expend. In all cases the mineral spirits cleaned faster and with less effort than the cold Pressroom Cleaner.

An additional test was conducted with mineral spirits and the Mirachem Pressroom cleaner. In this case, the full strength water-based cleaner cleaned faster and with less effort than the mineral spirits. The cost of the full strength Pressroom Cleaner, at about $10 per gallon, is higher than the cost of the mineral spirits by six times.

At that stage, IRTA suggested that R R Donnelley try to use water-based parts cleaners in place of the handwipe cleaning. The parts cleaner is a sink-on-a-drum which recirculates the cleaner and cleaning is performed with a brush in the sink. The cleaner is heated in the parts cleaner and could be used in diluted form so the cost would be lower. IRTA arranged for three vendors to provide parts cleaners to the company for a three-month testing period. After testing three different cleaning agents for a period, the workers that used the systems concluded that they favored the Mirachem parts cleaner. An advantage of the parts cleaners was the labor savings compared with handwiping and these costs are quantified below.

IRTA also suggested that the company test a spray cabinet that delivers high pressure spray to the parts. This increased mechanical action is generally a more effective cleaning method. IRTA arranged for a vendor to provide a spray cabinet to the machine shop for heavy duty cleaning for a three-month test period. The spray cabinet was used to clean all kinds of parts including rollers and ink pumps which are considered to be very difficult to clean. Because of the high pressure spray, the spray cabinet cleaned the parts more effectively than the handwiping and the parts cleaners. The spray cabinet also reduced the labor requirement substantially.

R R Donnelley has not purchased equipment at this stage. They are still evaluating the costs and trying to decide what mix of equipment would be optimal for their operation.

R R Donnelley used 5,463 gallons of mineral spirits in 1998 for press cleaning. The company estimates that 60 percent of the solvent used for press cleaning, or 3,278 gallons per year, can be replaced with water-based cleaning processes. This is the amount of solvent that is used for off-press cleaning. The remaining solvent, 2,185 gallons annually, is used in handwipe cleaning on the press itself. It is worth noting that the full strength Mirachem Pressroom cleaner performed well for cleaning on-press at the San Bernardino Sun where soy ink is used and might be expected to perform well at R R Donnelley where solvent based ink is used. The drawback to using this cleaner is its high cost at full strength.

R R Donnelley has six presses. The cost analysis below compares the cost of using different combinations of cleaning systems. The baseline is handwiping with mineral spirits. The scenarios that were analyzed are:
o Water-based cleaner with six spray cabinets
o Water-based cleaner with three spray cabinets
o Water-based cleaner with three spray cabinets with lower cost waste disposal and less frequent changeout
o Water-based cleaner with three spray cabinets and three immersion parts cleaners
o Water-based cleaner with one spray cabinet and six immersion parts cleaners

In 1998, R R Donnelley purchased 5,463 gallons of mineral spirits. at a cost of $1.67 per gallon. The water-based cleaning systems could replace 60 percent of the mineral spirits or 3,278 gallons. The total annual cost of purchasing the solvent for off-press cleaning was $5,474.

R R Donnelley's largest parts are 68 inches by 21 inches by 21 inches. A distributor quoted a price for an Anderson VSW-3000 spray cabinet which is large enough to clean these parts at $20,983. The Mirachem PW-50 immersion parts cleaners are priced at about $2,000. It was assumed that R R Donnelley would pay cash for the cleaning systems. The annual capital cost was determined by dividing the total capital cost by the equipment lifetime which was estimated at 10 years. The annual capital cost of a spray cabinet is $2,098; the annual capital cost of a parts cleaner is $200.

The water-based cleaner used in the spray cabinets is priced at $9.00 per gallon. The cleaner is used in a 10 percent solution in a spray cabinet. Each spray cabinet has a capacity of 150 gallons and the cleaning agent would have to be replaced monthly. Thus 15 gallons of water-based cleaner would be used in each spray cabinet each month. The cost of the water-based cleaner for one spray cabinet would be $1,620 annually; the cost of the cleaner for three spray cabinets would be $4,860 annually; and the cost of cleaner for six spray cabinets would be $9,720 annually. In the third scenario, a two-month changeout frequency is assumed and the annual cost of cleaning agent for each spray cabinet is $810.

The immersion parts cleaners would each have a 50 gallon capacity and the water-based cleaner would be used in a 25 percent solution. The cleaning solution would also be replaced monthly. Again the cost of the water-based cleaner is placed at $9 per gallon for the concentrate. The annual cost of the cleaning agent for each parts cleaner amounts to $1,350.

The number of hours spent cleaning off-press components with the mineral spirits is estimated by R R Donnelley at 5,682 annually. The labor rate is $9.50 per hour and the total labor cost for the solvent cleaning amounts to $53,979.

Spray cabinets are automated and the worker must only load and remove the parts. The labor hours spent cleaning with the spray cabinets are assumed to be one-tenth that of the labor hours spent handwiping. The labor hours required for using the immersion parts cleaners is assumed to be three-fourths of the labor hours used in hand cleaning. This estimate is based upon the times for cleaning during the side-by-side testing. In the scenario with three spray cabinets and three immersion parts cleaners, the workload is split evenly between the spray cabinets and the parts cleaners. In the scenario with one spray cabinet and six immersion parts cleaners, the spray cabinet would handle 25 percent of the workload.

Approximately one-half of the solvent used is disposed of as a hazardous waste. R R Donnelley pays $3 per gallon for segregated solvent waste. The annual cost of waste disposal for the solvent amounts to $4,917.

For the water-based system, each bath would be replaced monthly. At a capacity of 150 gallons each, each spray cabinet would generate 1,800 gallons of waste annually. At a capacity of 50 gallons, each immersion parts cleaner would generate 600 gallons of waste annually. In all but one of the scenarios, the waste was considered to be RCRA hazardous waste. The cost for disposal, assuming it is hazardous waste, is placed at $2 per gallon. The cost for disposal of the spent cleaner from one spray cabinet is $3,600 annually. The cost for disposal of the spent cleaner from one of the parts cleaners is $1,200 per year.

In the fourth scenario, the waste was considered to be non-RCRA waste. Since the wastestream could not be analyzed, the waste could be either RCRA hazardous or non-RCRA waste. This scenario was included so that all possibilities could be considered. Because the changeout frequency is half that of the other scenarios, the amount of waste generated in this scenario is also half. In this scenario, the cost for disposal of the waste was assumed to be lower, at 87 cents per gallon. The cost for disposal of the spent cleaner from one spray cabinet under this scenario is $783 annually.

There is an increased electricity cost from the use of the water-based cleaning systems. The spray cabinets have a 10 HP pump and a 24 kW heater. It is assumed that the pump operates one-fourth of the time and that the heater cycles on and off one-fourth of the time. R R Donnelley operates three 8-hour shifts for 250 days per year. Thus the heater and pump in the spray cabinets are operating 6 hours per day. The electrical use for each spray cabinet is 47,250 kWh annually. R R Donnelley's electrical rate is 10 cents per kWh. On this basis, the electrical cost for each spray cabinet is $4,725 annually. Each parts cleaner typically adds $10 per month or $120 per year to a company's electric bill.

R R Donnelley pays SCAQMD emission fees of $292.80 per ton for the mineral spirits solvents. Assuming that half the solvent purchased is waste and half is emitted, the annual emission fee amounts to $1,560.

Table 2-19 compares the costs of the six scenarios.

o Scenario 1-- baseline mineral spirits
o Scenario 2-- six spray cabinets
o Scenario 3--three spray cabinets
o Scenario 4-- three spray cabinets with less frequent and lower cost disposal
o Scenario 5-- three spray cabinets and three parts cleaners.
o Scenario 6-- one spray cabinet and 6 parts cleaners.

The values of the table indicate that the cost of three of the water-based cleaning system scenarios are lower cost than the mineral spirits scenario and two are higher. R R Donnelley could purchase three spray cabinets as alternatives to the mineral spirits cleaning. This would reduce the cost by 36 percent. If R R Donnelley purchased six spray cabinets, the cost would be higher than the use of mineral spirits by about 18 percent. The cost to R R Donnelley of purchasing three spray cabinets and three parts cleaners is only slightly higher, by about 3 percent, than continuing to use mineral spirits.

The emission reduction that could be achieved through substitution of the water-based cleaning systems amounts to 5.3 tons of VOC per year.

2.7. Cleaning of Ink Application Equipment--Screen Printing

Screen printing operations are widely varied. The process is used for printing on T-shirts or printing the serial numbers on military hardware. The inks used for screen printing can be solventborne or waterborne depending on the application. At the end of a print run after the ink is squeezed through the screen, the screens are cleaned. Inks with high resistance to heat and solvents tend to be the most difficult to clean.

During the project, IRTA worked with one company, Nelson Name Plate, that has a screen printing operation. The details of the conversion are described below.

2.7.1. Nelson Name Plate

Nelson has a screen printing operation that is part of the process of manufacturing the name plates. The company cleans about 70 screens per day. In the first cleaning operation, workers apply ink to screens and a squeegee is used to press the ink through the screen. This leaves an image behind on the substrate. The excess ink that remains on the screen must then be removed. The cleaner that is used currently is CCI-PW Screen Wash. The cleaner that is used to clean the screens must not damage the water soluble blockout that lies between the screen and the ink or the stencil itself; it must only remove the excess ink. A picture of Nelson's screen printing operation is shown in Figure 2-5.

Two water-based cleaners were tested as alternatives. One of these is a Mirachem formulation called Pressroom Cleaner; the other, made by W.R. Grace, is called Daraclean 257. Pure acetone and a 50 percent blend of acetone and the currently used cleaner were also tested.

The two water-based cleaners failed to remove all of the excess ink. The pure acetone attacked the water soluble blockout. The most effective alternative was a blend of the

cleaner currently being used with 50 percent acetone. Nelson decided to convert to this product.

About 1,275 gallons of the original high VOC content cleaner at 4.88 pounds per gallon were used by Nelson for screen cleaning each year. The cost of this cleaner was $16.34 per gallon. The total annual cost for the cleaner was $20,834. After the conversion to the lower VOC content solvent which has a VOC content of 2.44 pounds per gallon, the amount of cleaner used remained the same. Nelson reduced their cost in the conversion; the new cleaner had a lower cost of $10.65 per gallon. The total cost for purchases of the new cleaning agent amounts to $13,579 annually.

Disposal fees have remained the same for this operation. When the high VOC solvent was used, the rags used for the wipe cleaning were sent to an industrial laundry and returned to Nelson for reuse. The same number of rags were used after conversion to the lower VOC cleaner and the disposal was handled in the same way.

Nelson pays fees to the SCAQMD for VOC emissions. The current fee amounts to $292.80 per ton of VOC. When Nelson used the 4.88 pound per gallon VOC cleaner, the emissions were 6,222 pounds or 3.11 tons per year. The emissions fee was about $911 annually. After conversion to the 2.44 pound per gallon VOC cleaner, the emissions were cut in half, to 3,111 pounds or 1.56 tons annually. The emission fee, at $455 annually, was also cut in half.

Table 2-20 summarizes and compares the costs for the high and low VOC solvent cleaning operations. The conversion to the lower VOC content cleaner reduced Nelson's costs by 35 percent. Nelson realized a cost savings of about $5,000 per ton of VOC emitted.

After a print job is completed, the screens are treated with a VOC solvent to remove dried ink and sprayed at high pressure with a water-based cleaner and recycled for reuse. At this stage, the cleaner is removing any blockout and ghost image that may remain. Figure 2-6 shows the spray cleaning operation. Many other companies that do screen printing use VOC solvents for this cleaning operation. One commonly used solvent is d-limonene, a terpene.

About five years ago, Nelson used a solvent cleaning process for recycling the screens. Purchase of the solvent for this purpose amounted to about $1,000 per month. VOC emissions from this process were about 200 pounds per month or about 1.2 tons annually. To reduce VOC emissions, Nelson converted to the water-based process they use currently. In this process, all of the screens are first treated with a VOC cleaner with a VOC content of about 5 pounds per gallon to remove dried ink. Nelson uses about 15 gallons per month of this cleaner; emissions amount to 75 pounds per month. Then the screens are treated with one of four alkaline water-based chemicals depending on the ink. These treatment chemicals contain no VOC. The cost of the chemicals amounts to about $1,000 per month. Then the screens are rinsed with high pressure tap water. The costs of using the VOC solvents and the water-based process are roughly equivalent and the VOC emissions have been reduced by 125 pounds per month.

III. RULE 1122--SOLVENT DEGREASERS

 SCAQMD Rule 1122 regulates VOC solvents used in batch loaded and conveyorized cold cleaners and vapor degreasers. In July, 1997, the rule was substantially amended. It set new standards for batch loaded cold cleaners (BLCCs), open top vapor degreasers, conveyorized vapor degreasers and conveyorized cold cleaners

3.1 Batch Loaded Cold Cleaners and Conveyorized Cold Cleaners

The rule amendment set standards for BLCCs and conveyorized cold cleaning units prior to January 1, 1999. After January 1, 1999, solvents used in BLCCs were required to have a VOC content of 50 grams per liter or less or be used in an airless or air-tight cleaning system. After January 1, 1999, conveyorized cold cleaners were required to use cleaning materials that have a VOC content of 50 grams per liter or less. An exemption was provided in Rule 1122 for certain types of operations. It allowed the continued use of unheated BLCCs until January 1, 2003, with open top surface areas less than 1.0 square foot or with a capacity of less than 2 gallons for cleaning particular components like optics or electronics applications. It restricted the solvent usage to less than 5 gallons per month and it specified operating requirements for the cleaning units.

Historically, BLCCs used a variety of VOC solvents including mineral spirits, methyl ethyl ketone, methyl isobutyl ketone, toluene, xylene, isopropyl alcohol, glycol ethers and various blends of some of these solvents. Some conveyorized cold cleaners commonly used terpenes, glycol ethers, isopropyl alcohol or blends of water and these solvents. These latter solvents were generally adopted as alternatives to ozone depleting or chlorinated toxic solvents that were used in vapor degreasers (see discussion on vapor degreasers below). Other conveyorized cold cleaners used mineral spirits or some of the other solvents commonly used in BLCCs. By January 1, 1999, operations using these VOC solvents had to comply with the rule requirements.

This section features three companies that converted BLCCs using VOC solvents to water-based cleaning systems. One of these companies, S&H Machine, Inc., is a machine shop that employed BLCCs using mineral spirits. This company made a conversion to water-based cleaners to comply with the provision of Rule 1122 that became effective on January 1, 1999. S&H Machine is representative of hundreds of machine shops in the South Coast Basin that have similar operations and needs.

The second company, Hydro-Aire, is an aerospace subcontractor. This company used four BLCCs with mineral spirits in their machine shop and repair and overhaul areas. This company converted to water-based parts cleaners to satisfy the Rule 1122 provision. Although Hydro-Aire cleans aerospace parts, they are not covered by the cleaning provisions of Rule 1124 which apply only to handwipe operations. Some of the components cleaned by Hydro-Aire, however, may fall under the exemption in the rule that is referred to above.

The third company, Litton Guidance & Control Systems, converted from a variety of VOC solvents used for precision cleaning of optics to water-based cleaning systems. Litton's operations fall under the exemption provision in Rule 1122 for BLCCs used for cleaning optics. Litton elected to make the conversion even though the company could have exercised their exemption until January 1, 2003.

 3.1.1. S&H Machine, Inc.

This company, located in Burbank, California, is a small machine shop that machines parts for aerospace applications. S&H Machine is typical, in terms of size and processes, of numerous machine shops located in the Basin.

For many years, the firm used mineral spirits at 21 separate stations for cleaning their parts as they were machined. The purpose of the cleaning is to remove oil and chips from the machined parts so they can be inspected. The workers would machine the parts, dip them in a coffee can containing mineral spirits, hold them over the machine and blow them off with compressed air. The oil used in the machines required dilution with mineral spirits and the excess mineral spirits on the parts during blow-off satisfied this requirement. S&H Machine also had two larger 5-gallon batch loaded cold cleaners that contained mineral spirits. These cleaning units were used to perform final cleaning of the parts. Parts were placed in the unit, cleaned individually one-by-one and blown off with shop air by the workers.

S&H Machine purchased 17 55-gallon drums of mineral spirits annually at a cost of $148.50 per drum. The total annual cost of the mineral spirits was $2,525. The shop had no disposal cost because, when the parts were blown off with air, the mineral spirits on them was simply added to the machines for dilution of the coolant.

SCAQMD Rule 1122 required firms to use cleaners with 50 grams per liter VOC or less or to convert to an airless or air-tight degreaser. S&H decided to convert to water-based cleaners and IRTA worked with the facility to determine what their needs would be. IRTA and S&H tried two different approaches. First, a neutral water-based cleaner that would not damage the workers' hands was used in the same coffee cans that that contained the mineral spirits. The problem with this arrangement was that the water cleaner that was chosen, Daraclean 236 which is made by W.R. Grace, was designed to reject oil. The oil would float on the top of the liquid and the parts were recontaminated with the oil when they were removed from the coffee cans.

Second, IRTA arranged for a water-based 30-gallon plastic sink-on-a-drum made by Gray Mills to be installed in the facility to see if this type of cleaning system could replace the coffee cans of mineral spirits. This approach worked well. The company ended up purchasing this system for $900; it is shown in Figure 3-1. S&H also purchased seven other smaller Gray Mills 15-gallon metal sink-on-a-drum parts cleaners second hand; these cleaning units together cost the shop $900. One of the units is shown in Figure 3-2.

S&H Machine wanted to purchase a larger cleaning system to replace the two 5-gallon batch loaded cold cleaning units used for final cleaning. The aim was to minimize the labor spent during final cleaning of the parts. After testing in a small tabletop unit loaned to the facility, S&H Machine decided to purchase an ultrasonic cleaning system made by Western Sonics for $19,008. A picture of this system is shown in Figure 3-3.

S&H Machine paid cash for all of the new cleaning equipment. Assuming the equipment has a useful life of 10 years, the annual cost of the ultrasonic system and the parts cleaners is $2,081.

IRTA and S&H experimented with water-based cleaners to find the optimal one for the shop. A neutral cleaning agent would have to be used in the parts cleaners where the workers have hand contact. IRTA and S&H Machine identified a neutral cleaning agent that was capable of removing the honing oil from the assemblies in the ultrasonic cleaning unit and was also effective in the parts cleaners. This cleaner is made by Applied Cleaning

Technologies (ACT). The facility can use this cleaner in the parts cleaners and the ultrasonic system. When the ultrasonic wash bath is spent, it can be used as makeup for the parts cleaners which can tolerate a higher contamination level. This will minimize disposal costs for S&H Machine. The cleaning agent, called Scrub Tub, has been certified as a Clean Air Solvent by SCAQMD.

The original Gray Mills plastic unit has operated for about seven months and has not required changeout. Assuming that this unit and the other seven parts cleaners require changeout twice per year, the amount of cleaner required is 270 gallons. S&H Machine is using a 10 percent concentration of cleaning agent in the cleaning units. Thus 27 gallons at a cost of $8.50 per gallon are required for changeout. The spent cleaner and rinse water from the 30-gallon ultrasonic unit can be used as makeup in the parts cleaners so no additional cleaner needs to be used for this purpose. The total annual cost of the water based cleaner for use in the parts cleaners is $230.

The 30-gallon ultrasonic cleaning unit uses the same cleaner also at a 10 percent concentration. Assuming it requires changeout more often, every two months, the cost of the cleaning agent for changeout is $153 per year. An additional 10 percent of the cleaning agent is needed as makeup. The total cost of the cleaning agent for use in the ultrasonic system is $168 annually.

The company also uses a rust inhibitor in the rinse bath of the ultrasonic system. The price of the rust inhibitor is $11 per gallon. A concentration of three percent is required in the 30-gallon bath. This amounts to about $10 per bath. The bath must be changed out every two months. An additional 10 percent of the rust inhibitor is required as makeup. The total annual cost of the rust inhibitor is about $66.

S&H Machine workers used compressed air to blow off the excess solvent after cleaning back into the machines. With the conversion to water-based cleaners, no solvent was being added to the machines. The coolant still requires some solvent for dilution and the amount is estimated at one drum every three months. At a price of $148.50 per drum, the total annual cost for the dilution solvent amounts to $594.

The spent water-based cleaner will require disposal. It is estimated that S&H Machine will generate about four drums of waste per year. The waste has not yet been analyzed but it is likely that it will be non-RCRA hazardous waste. The waste hauler that S&H plans to use for disposal charges $46.75 per drum for disposal of non-RCRA aqueous waste. The total cost for disposal is $187 per year.

S&H Machine estimates that the labor for cleaning with the mineral spirits and water-based cleaner is about the same at the machine stations. For the final cleaning area where the 5-gallon cleaning units were used, workers cleaned for about one hour per day. The ultrasonic cleaning unit is automated and the parts are cleaned in baskets instead of individually. The workers still have to blow off the parts. It is estimated that the ultrasonic cleaning system reduces the labor requirement to 40 percent of that required for the solvent cleaning. At a labor rate of $15 per hour, the labor cost for cleaning with the mineral spirits was $3,900 per year. With the ultrasonic system, the labor cost is $1,560 per year.

The electrical cost increases with water-based cleaning. S&H Machine cannot estimate the increase in the electrical use since installation of the new cleaning units; they have added other equipment to the shop that has also increased the electrical use. The electrical cost of the large Gray Mills unit is estimated at $10 per month or $120 per year. The electrical cost of each of the seven smaller parts cleaners is estimated at half the amount or a total of $420 per year. The total electrical cost for the parts cleaners is $540 annually.

The ultrasonic system has 1,200 watts of ultrasonic power in the wash tank. Assuming the system is used one-half hour per day and an electricity cost of 12 cents per kwh, the electrical cost from using the ultrasonics is about $19 per year. The wash and rinse bath are heated and each of the heaters is 2,250 watts. These heaters are left on at night during the week and are turned off on the weekend. They are likely to be cycling on about one-fourth of the time they are operating. Again, assuming an electricity cost of 12 cents per kwh, the cost of running the heaters is $842 annually. The total electrical cost of the ultrasonic system is $861 per year.

Table 3-1 shows the costs to S&H Machine for using mineral spirits and water-based cleaners. The figures show that the conversion to the water-based cleaning systems resulted in a small savings of about 3 percent. Although S&H Machine had to purchase new equipment and increased their electric load, this cost was more than offset by the savings in labor from use of the ultrasonic unit and the lower cleaning agent cost.

S&H Machine reduced their mineral spirits purchases and emissions by 13 drums per year. The company originally used 17 drums of solvent per year which was eliminated with the adoption of the water-based cleaning systems. The coolant used in the machines still requires some dilution with solvent and four drums of solvent are required for this purpose. The net reduction in use and emissions for S&H Machine is 13 drums of mineral spirits. Assuming a solvent density of 6.5 pounds per gallon, this amounts to about 4,648 pounds or 2.3 tons of VOC solvent annually. The use of the water-based cleaner reduced S&H Machine's cost by $57 per ton of VOC emitted.

A stand-alone case study for S&H Machine is provided in Appendix B.

3.1.2. Hydro-Aire

Hydro-Aire is an aerospace subcontractor that has a variety of cleaning operations. The company used four batch loaded cold cleaners with mineral spirits in their facility. Two of these, with a 3-gallon capacity, were in the machine shop. The captive machine shop used the cleaning units to clean oil and chips from the parts so they can be inspected. The other two cleaning units, with a 5-gallon capacity, are in the repair and overhaul area. These systems are used to clean parts that have come from the field. They are often covered with grease and grime accumulated from years of use. Total annual VOC emissions from these cleaning units are estimated at about 450 pounds or 0.23 tons annually.

IRTA worked with Hydro-Aire to test water-based sink-on-a-drum parts cleaners in the applications where the mineral spirits batch loaded cold cleaners were used. Hydro-Aire purchased four Kleen-Tech parts cleaners at a cost of $1,200 each. A picture of one of these units is shown in Figure 3-4. Assuming a 10-year life for the units and that Hydro-Aire paid for them in cash, the annual cost for all four units is $480.

The cost of the mineral spirits used in the batch loaded cold cleaners was $1.96 per gallon. The 3-gallon capacity units required changeout every three weeks. The 5-gallon units required changeout every six weeks. The cost of the mineral spirits for all four units amounted to $374 annually.

The parts cleaners use a water-based cleaner priced at $12.90 per gallon. The concentration in the 30-gallon cleaning units is 20 percent. Assuming that the cleaner is changed out every six months and assuming that 10 percent cleaner is required for make-up, the cost of the cleaning agent is $681 annually.

A company provides servicing/maintenance on the water-based cleaning units for Hydro-Aire. The company spends about one-half hour every two weeks for servicing. Twice each year, when the units require changeout, the servicing will take two hours. At a cost of $45 per hour, the servicing cost amounts to $720 annually.

The mineral spirits disposal cost amounted to $75 per drum. For the four batch loaded cold cleaners the total annual disposal cost was $300.

The spent water-based cleaner must be sent for disposal. Assuming the spent cleaner is non-RCRA hazardous waste, the cost for disposal is $46.75 per drum. The total annual disposal cost is $47.

There was no electrical cost for the batch loaded cold cleaners because the cleaning units had no pump and the solvent was not heated. The electrical cost for a sink-on-a-drum parts cleaner is estimated at $120 per year. For the four parts cleaners, the total electrical cost is $480 annually.

A Hydro-Aire representative estimates that the labor required for the mineral spirits and the water-based cleaning is roughly equivalent.

Table 3-2 compares the annual costs to Hydro-Aire for using the mineral spirits in batch loaded cold cleaners and using the water-based cleaners in the parts cleaners. The cost of using the water-based cleaners is three-and-a-half times more than the cost of using the mineral spirits.

Hydro-Aire reduced their VOC emissions from the BLCCs by 0.23 tons per year at a cost increase of $1,672 per year. The conversion resulted in a cost of $7,300 per ton of VOC reduced.

A stand-alone case study on this operation and another Hydro-Aire cleaning operation is presented in Appendix B.

It is worth noting that in the previous case study, S&H Machine, the costs of using water-based cleaners and mineral spirits were comparable. In the case of Hydro-Aire, the cost of using water-based cleaners is much higher than the cost of using mineral spirits. Hydro-Aire opted for servicing of the water-based cleaner whereas S&H Machine did not. S&H machine also reduced their cleaner cost and their labor cost substantially through the conversion. Hydro-Aire increased the cleaner cost and had no change in labor cost. In addition, S&H Machine, a small machine ship, paid particular attention to minimizing their cost during the transition.

3.1.3. Litton Guidance & Control Systems

Litton manufactures laser guidance systems for commercial and military aerospace applications including spacecraft and aircraft missiles. A picture of one of the components of this system is shown in Figure 3-5. The high precision parts are lapped and polished and blocking materials are used to hold the parts in place during these operations. The parts are cleaned in several steps of the process to remove the lapping, polishing and blocking compounds.

In the past, Litton relied heavily on CFC-113 and TCA for cleaning the parts. Several years ago, Litton initiated projects to find alternatives. They converted primarily to VOC solvents and some water-based cleaning processes. At that stage, the company's operations were classified as BLCCs using VOC solvents and were covered by Rule 1122. Litton did not have to make a conversion away from the VOC solvents in 1999 because they qualified for an exemption, (k)(1)(c), that extended the deadline until 2003. Even so, the company decided they wanted to convert away from the VOC solvents in 1998 and they again began working on non-VOC alternatives. By January, 1999, Litton had reduced their use and emissions of VOC solvents by 16,000 pounds or 8 tons annually.

In the frame manufacturing operation, Litton used n-methyl pyrollidone (NMP) to clean wax which was used to plug the frame bores to prevent lapping compound from intruding. The company eliminated this cleaning step by using plugs with O-rings to block the frame bores acting as a physical barrier to the lapping compound. In another step, epoxy was used to bond the frames to holding fixtures during lapping and polishing. NMP was used to remove the epoxy. Hot air at a temperature of 200 degrees F is now used to separate the frame from the fixture. The thermal expansion differences between the glass frame, metal fixture and epoxy causes the debonding. A picture of this operation is shown in Figure 3-6.

In another operation, the substrate operation, pitch was used to hold the mirror substrates to mounting blocks during lapping and polishing. NMP, Bioact 280, a terpene-based solvent and small amounts of methanol and methylene chloride were used for deblocking and cleaning. Litton has substituted a thermoplastic for the pitch in the bonding operation. Acetone is currently used to dissolve most of the thermoplastic; this cleaning step is followed by a soak in an Armakleen detergent that is a certified Clean Air Solvent. Acetone was also used in the past for the cleaning.

In the prism operation, wax is used to bond the prisms to mounting blocks for lapping and polishing. A terpene product called Opticlear was used to dissolve the wax and clean the parts. This product has been replaced by Daraclean 121, a water-based cleaner. The new cleaning operation is shown in Figure 3-7.

Litton used 10 drums of NMP per year in their process in the past. The cost of the NMP was $450 per drum. The total annual cost of the NMP was $4,500. Fourteen drums of Bioact 280 were used each year at a cost of $550 per drum. The total cost of using the Bioact was $7,700 per year. Fourteen drums of Opticlear were also used each year at a cost of $1,695 per drum. The total cost of using the Opticlear was $23,730 annually. The
Figure 3-5 Laser Gyro Showing Optical Components at Litton

cost of the methylene chloride and the methanol amounted to about $200 per year. The total yearly cost for purchasing the VOC solvents was $36,130.

The new operations involve the use of Daraclean 121 and an Armakleen cleaning agent. Litton estimates that three drums of Daraclean 121 at a cost of $850 per drum will be required. Two drums of the Armakleen detergent at $105 per drum will also be required. The total cost of the two water-based cleaners amounts to $2,760 annually.

Litton substituted thermoplastic for pitch in the bonding operation. The thermoplastic, at a cost of $12,000 annually, is much more expensive than the pitch which carried a cost of about $2,000 annually.

Disposal costs for the Bioact 280 were $1,890 per year. Disposal costs of the Opticlear was also $1,890 per year. The disposal cost for the NMP was $1,350 per year. The total disposal cost for the solvents amounted to $5,130 annually.

The disposal cost for the Daraclean 121 is $405 per year. The Armakleen detergent does not have a disposal cost. Litton is exploring whether the thermoplastic can be recycled.

Litton estimates that the electrical cost and the labor cost remain the same with the old and new operations.

Table 3-3 shows the cost comparison for the VOC solvents and the water-based cleaners. By making the conversions to not cleaning and to water-based cleaning, Litton reduced their emissions by about 8 tons per year. They also reduced their costs by about 65 percent. The conversions resulted in a cost savings of $3,512 per ton.

A stand-alone case study of Litton's conversion is provided in Appendix B.

3.2. Vapor Degreasing

Vapor degreasing is generally performed with halogenated solvents because the process cannot be used with solvents that have a flash point. The solvents that have been used historically and are still used in vapor degreasing include 1,1,1-trichloroethane (TCA), 1,1,2-trichloro-1,2,2-trifluoroethane (CFC-113), perchloroethylene (PERC), methylene chloride (METH) and trichloroethylene (TCE). Other new halogenated solvents that have been developed more recently include HCFC-141b, HCFC-123, HCFC-225, HFC-43-10, HFEs, 1,2-trans-dichloroethylene (DCE) and n-propyl bromide (NPB). In many cases, these halogenated solvents are combined with other solvents like alcohols or acetone for particular cleaning applications.

Of the solvents listed above that are used in vapor degreasing, most are exempt from VOC regulations. Those that are classified as VOCs include TCE, DCE and NPB. DCE is not used alone because it has a flash point. It is generally combined with HFC-43-10 and the HFEs in blends that do not have flash points.

The Rule 1122 vapor degreasing provisions apply to VOC solvents and solvent blends that have 50 grams per liter or more VOC content. Vapor degreasers using nPB would be covered by the rule because the chemical is a VOC. Vapor degreasers containing blends of HFC-43-10 and DCE or HFEs and DCE would be covered by the rule; generally these blends contain more than 50 percent DCE. Other combinations of the exempt solvents listed above and alcohols of various kinds would be covered by the rule if the VOC content exceeded 50 grams per liter.

Rule 1122 provides an exemption for the solvents covered by the Halogenated Solvent Degreasing National Emission Standards for Hazardous Air Pollutants (NESHAP). The solvents that are covered by the NESHAP include chloroform, carbon tetrachloride, TCA, PERC, TCE and METH. Thus vapor degreasing operations involving these solvents are not covered by Rule 1122. The NESHAP was effective on new sources in 1994 and became effective on existing sources in December, 1997. In general, it requires companies to use tighter vapor degreasers. The District is responsible for implementing the NESHAP in the South Coast Basin.

In January, 1996, production of certain ozone depleting solvents was banned worldwide. The solvents that were affected by this ban include TCA and CFC-113. Inventory of the solvents remained and many companies continued to use them. There is also a congressional tax on the two solvents which makes them very expensive. At this stage, most companies have converted away from the solvents. Under EPA ozone depletion regulations, HCFC-141b is no longer allowed to be used for vapor degreasing. HCFC-123 and HCFC-225 will be banned at a later date because they also contribute to ozone depletion. There are a few companies using HCFC-225 and its blends with VOC solvents that might be covered by Rule 1122. HCFC-123 has been taken off the market because it caused liver problems in workers.

Many of the companies that used TCA and CFC-113 converted to VOC solvents. Litton Guidance & Control Systems which is highlighted in the last subsection is an example of such a facility. Other companies converted to PERC. These companies are subject to the NESHAP and may be forced to reduce their emissions further because of SCAQMD Rule 1402 which covers existing sources of toxics. Still other companies have converted to the HFC or HFEs, generally in blends with VOC solvents. By far the largest number of companies have converted to water-based cleaners and such processes are appropriate for virtually all vapor degreasing operations that still use solvents.

According to the District staff report that was written for the most recent Rule 1122 amendment, the inventory for vapor degreasers and conveyorized cold cleaners was 2.8 tons of VOC per day in 1999. The District estimates that this will grow to 12.4 tons per day by 2010.

In this subsection, the focus is on companies that converted away from solvents in vapor degreasing operations. Eight operations that involve vapor degreasing are evaluated. Companies have converted away from vapor degreasing in general metal cleaning, printed circuit board defluxing and precision cleaning. Each of these categories is discussed in more detail below.

3.2.1. General Metal Vapor Degreasing

Four of the companies that are featured converted away from TCA or PERC in general metal vapor degreasing. These include California Electroplating Inc., a plating shop that used PERC, Hydro-Aire, an aerospace subcontractor that used TCA, Nelson Name Plate, a name plate manufacturer that used TCA and Barranca Diamond Products, a diamond saw blade manufacturer that had an expired PERC permit. Instead of converting to VOC solvents or blends of exempt and VOC solvents, all of these companies converted to water-based cleaning processes.

3.2.1.1. California Electroplating Inc.

California Electroplating is a small company with 19 employees located in Los Angeles. The company provides high volume chromium and nickel plating for zinc die cast, steel and brass parts. The facility has two plating lines. In the automated line, which is used for high volume jobs, the parts, which do not contain polishing compound, are cleaned in-line. In the hand line which is used for plating custom smaller volume jobs, the parts are contaminated with polishing or buffing compound.

California Electroplating used a PERC vapor degreaser to clean all of the parts. IRTA began working with the company to identify, test and implement a water-based cleaning system as an alternative to the PERC degreaser. After testing parts containing a buffing compound, a cleaning agent that is effective in removing the compound was selected. This cleaner is called Daraclean 236 and is certified by SCAQMD as a Clean Air Solvent. The company purchased an ultrasonic water-based cleaning system to clean the parts. A picture of the new ultrasonic system is shown in Figure 3-8.

 At this stage, California Electroplating has converted to a water-based process for the parts on the custom line which are cleaned on plating fixtures. The vapor degreaser is still used for the parts on the automated line. California Electroplating plans to install a hoist within the next few months so the parts from the automated line can be cleaned in baskets in the water-based system.

The company used about 100 gallons of PERC per month before the purchase of the water-based system; the use has been reduced to about 100 gallons every three months. At a cost of PERC of $6.75 per gallon, the PERC purchases amounted to $8,100 annually before the water system was installed and $2,700 after the installation. Waste generation was about 2 drums every three months; this has been reduced to about 2 drums every nine months. At a cost for disposal of $100 per drum, the cost before the water system installation for hazardous waste disposal was $800; the cost after the water system is $267.

The cost of the ultrasonic cleaning unit was $21,670 and the company paid cash. Assuming a ten-year equipment lifetime, the annual cost of the water cleaning unit is $2,167. The cost of the water-based cleaner is $500 per drum or about $9 per gallon. The 150 gallon cleaning unit requires an eight percent charge of the cleaning concentrate. Thus 12 gallons are required to charge the system. The bath is changed out every two months. The annual cost of the cleaning agent for charging the system is $648. Make-up cleaner is also required to replace evaporation and drag-out; this amounts to 10 percent of the charge. The total cost of the cleaner amounts to $713 per year. The cleaning agent does not require disposal; the plating facility treats thousands of gallons of wastewater each day and the bath is discharged to the treatment system every two months.

Frank Grana, one of the owners of California Electroplating, estimates that before the water system was installed, the labor used to clean in the vapor degreaser was three hours per day. The labor now, after the installation of the water system, is only about three hours per week. Mr. Grana also estimates that the labor used in the water-based cleaning is only 50 percent of the labor used in vapor degreasing. At a labor rate of $8 per hour, the labor cost for vapor degreasing before the conversion was $6,240. The labor cost for vapor degreasing after the conversion is $2,496. The labor cost for cleaning with the water-based system is $1,248 per year.

The ultrasonic cleaning unit contains 4,800 watts of power. Assuming that the ultrasonics are operating about four hours each day or 1,040 hours per year and that the cost of electrical power is 12 cents per kW, the electricity cost for running the ultrasonics is $599 per year. The unit is heated with a boiler. Mr. Grana estimates that the gas cost for the ultrasonic system amounts to $50 per month or $600 per year. Although the degreaser uses gas for heating as well, this cost is assumed to be negligible.

Table 3-4 shows the costs for using the PERC degreaser and the water-based cleaning system. Conversion to the water-based cleaning system reduced the costs by 29 percent. Although the company had to purchase a new water-based cleaning unit and there was an increase in utility costs, this is offset by the lower cost of the cleaning agent and a reduction in cleaning labor.

Because PERC is exempt from VOC regulations, the conversion did not achieve a VOC reduction. A reduction in toxic emissions of about 730 gallons or about 5 tons annually was achieved, however. The cost savings was $870 per ton.

A stand-alone case study of California Electroplating's conversion is provided in Appendix B.

3.2.1.2. Hydro-Aire--Vapor Degreaser Conversions

Hydro-Aire is a division of Crane located in Burbank. The company manufactures fuel and coolant pumps, hydraulic control components and braking control systems and is a Boeing subcontractor. The firm utilized four vapor degreasers that used TCA to clean their parts for many years. IRTA began working with the company in 1997 to assist them in converting to water-based cleaning systems.

Several different water-based cleaners were tested. Two were ultimately selected for the three areas in the plant where parts require cleaning. An ultrasonic system was necessary because Hydro-Aire's parts have blind holes that are difficult to clean. Three systems were specified and ordered from Western Sonics. All three systems have been built and installed recently at Hydro-Aire. The first system is in the deburring section; the second system is in the hone, lap and grind area; and the third system is in the non-destructive testing area. A picture of the system in the deburring area is shown in Figure 3-9.

The three ultrasonic cleaning units each have an ultrasonic wash bath, two ultrasonic rinse baths and a dryer. All three systems are also fully automated and robotic. One of the rinse baths is routed to a closed loop water recycling system. The cost of the three cleaning units was $274,844. The cost of the two closed loop water recycling systems that serve the three cleaning units was $30,000. Hydro-Aire paid cash for the systems. Assuming a 10-year equipment life for the ultrasonic units and the water treatment systems, the annualized capital cost was $30,484.

The four vapor degreasers operated by Hydro-Aire together used 1,400 gallons of TCA per year. At a cost of $86.60 per gallon, the total annual cost for purchasing TCA was $121,240. Each of the three water-based cleaning units has about a 30 gallon wash bath capacity. The cost of the water-based cleaner is $12.90 per gallon and a 10 percent concentration is required. Assuming that each bath is changed out once a month and that makeup cleaner amounts to 10 percent, the total cost to Hydro-Aire for the water-based cleaning concentrate is $1,533 annually.

Maintenance on the vapor degreasers required 120 hours per year at a worker labor rate of $12 per hour. The total cost for maintenance of the vapor degreasers was $1,440. A company provides the maintenance on the water-based cleaning systems for Hydro-Aire every two weeks at a cost of $45 per hour. Assuming two hours are required every maintenance period, the total cost of the maintenance is $2,340 annually. The deionized water for the rinses is obtained by installed beds. The cost of maintaining these beds is $1,000 annually.

The labor used for cleaning with the vapor degreasers in the deburring area was 5 minutes per batch cleaned. Assuming that 18,720 batches were cleaned and a labor rate of $15 per hour, the labor cost in the deburring area was $23,400. In the hone, lap and grind area, 12,220 batches were cleaned at 4 minutes per batch. The labor cost of cleaning in this area amounted to $12,220 annually. In the destructive testing area, 26,000 batches were cleaned at 4 minutes per batch. The labor cost for this area was $26,000. The total labor cost for cleaning with the vapor degreasers was $61,620 annually.

Hydro-Aire estimates that the labor cost has been reduced by about 60 percent in the deburring area and by about 50 percent in the hone, lap and grind and destructive testing areas because the new water cleaning equipment is automated. Assuming these reductions, the total labor cost for cleaning with the water-based cleaner is $28,470 annually.

The waste disposal cost of the TCA amounted to $2,000 annually. Assuming that the water-based cleaning baths are changed out once a month, 90 gallons of aqueous waste are generated each month. The total amount of aqueous waste generated is 1,080 gallons or about 20 drums per year. At a disposal cost of $50 per drum for non-RCRA hazardous waste, the waste disposal cost with the water-based cleaner amounts to $1,000 annually. The rinse baths that contain rust inhibitor also require changeout, in this case every two months. Assuming the same cost for disposal, the cost of disposing of the rust inhibited rinse baths is $500 annually. The total disposal cost for the water-based cleaning systems is $1,500 annually.

The electricity cost increases with water-based cleaning. Each of the three cleaning systems has a 4,000 watt or 4 kW heater in each of the three baths. The heaters come on at 3:00 AM and go off at 5:00 PM each day. Assuming that the heater cycles on about one-fourth of the time, the electricity use for heating for the three cleaning units is 32,760 kW. At an electricity cost of 12 cents per kW, the cost of heating the units is $3,931. The ultrasonics in each system, at 1.6 kW each, operate for about five hours per day. The electricity usage is about 6,240 kW annually. At a cost of 12 cents per kW, the annual cost of using the ultrasonics is $749. The dryer, which has a 5 kW heater in each unit, operates about 3 hours per day. The electricity requirement is 11,700 kW annually. Again, at a cost of electricity of 12 cents per kW, the annual cost of operating the dryer is $1,404. The total electricity cost for the three water cleaning systems amounts to $6,084 per year.

Hydro-Aire was required to have permits and do recordkeeping and reporting when the TCA was used. Permit fees amounted to $6,000 per year. Emission fees were $4,000 per year. The cost to Hydro-Aire for recordkeeping is estimated at $2,300. The total cost of the regulatory compliance amounted to $12,300. The water-based cleaning units do not require permits, emission fee payments or recordkeeping.

Table 3-5 summarizes and compares the annual costs of the solvent degreasers and the new water cleaning systems. The water cleaning systems carried a high capital cost and the electrical cost also increased. These costs are offset by the much lower cleaner cost, lower labor costs and no regulatory fees. The annual costs were reduced by 64 percent through the conversion.

Hydro-Aire reduced their emissions of TCA by an estimated 1,120 gallons or 6.16 tons per year. Although TCA is not a VOC, the company could have converted to a VOC solvent for their cleaning. The cost savings to Hydro-Aire that was achieved through the conversion was $20,648 per ton of emission reduction.

Hydro-Aire's conversion is also described in a stand-alone case study in Appendix B.

3.2.1.3. Nelson Name Plate

Nelson Name Plate is a small firm with 200 employees in Los Angeles. The company manufactures name plates and membrane switches. Aluminum, stainless steel and brass stock for the name plates are shipped to the facility with a protective film of oil. Historically, Nelson used a TCA degreaser to remove the oil from the metal. About 1,000 sheets were cleaned in the vapor degreaser each day.

TCA production was banned on January 1, 1996 and the company needed an alternative that would not increase their VOC emissions. In 1995 and 1996, IRTA worked with the company to identify, test and implement a water-based cleaning alternative. After substantial testing, Nelson purchased a conveyorized cleaning system from Trek Industries and began using a water-based cleaner made by W.R. Grace. Since then, Nelson identified a new cleaner, also made by W.R. Grace, that is more effective for cleaning the parts.

The water-based cleaning system is shown in Figure 3-10. The capital cost of the system amounted to $128,000 and Nelson paid cash for it. Assuming a 10-year lifetime for the equipment, the annual cost was $12,800. For the firm to continue using the degreaser, an upgrade would have been required to comply with the NESHAP. Because Nelson converted before the NESHAP became effective on existing facilities, this cost is not factored into the analysis.

Before the conversion, Nelson was using about 400 gallons per month or 4,800 gallons per year of TCA. The cost of TCA, at the time, was $16.15 per gallon. This leads to an annual TCA cost of $77,520. The cost for disposal was very low, perhaps $200 per year.

The TCA is recycled so a premium is paid for the used TCA. Firms are charged for the contaminants but this often compensates for the payment for the sludge disposal.

Nelson is using the new water-based cleaner at a concentration of about 10 percent. Approximately 40 gallons of cleaner per month or 480 gallons per year are used by Nelson. At a cost of $8.24 per gallon, the annual cost of cleaner is $3,955. The wash and rinse baths are changed out about every two months. Nelson has a wastewater treatment system for their anodizing operation and the company has permission to discharge the water from the cleaning system. Nelson estimates that the disposal cost is about $100 per month or $1,200 per year.

Nelson operated the vapor degreaser for two shifts or 16 hours per day. A worker was needed to process the parts during operation. At a labor rate of $10 per hour, the labor cost for operating the vapor degreaser amounted to $41,600 annually. The water-based cleaning system reduced the labor hours to 10 per day because it is an automated machine. The annual labor cost with the water-based cleaning system is $26,000.

The electrical cost has increased since purchase of the new water-based cleaning system. Nelson estimates the increase at $100 per month or $1,200 per year.

Nelson paid emission fees to the SCAQMD for the TCA. Of the 4,800 gallons of TCA used each month, it is estimated that 75 percent accounted for air emissions and 25 percent left the facility as hazardous waste. Thus emission of TCA were about 3,600 gallons or 39,600 pounds per year. The SCAQMD charged three cents per pound for TCA emissions at the time. The total emission fee for Nelson amounted to $1,188 annually.

The costs are summarized in Table 3-6 below. The annual cost of using the vapor degreaser, at about $120,000, is about 2.7 times more than the annual cost of using the water-based cleaner. Nelson reduced the cost of cleaning by about 63 percent by converting to a water-based cleaning system.

Nelson's TCA emissions amounted to 19.8 tons per year. Although TCA is not a VOC, the company avoided a conversion to VOC solvents for cleaning by converting to the water-based cleaner. Nelson's savings in making the conversion amounted to $3,806 per ton.

A stand-alone case study for Nelson is provided in Appendix B.

3.2.1.4. Barranca Diamond Products

Dean Delahaut purchased Barranca Diamond Products, a small company with seven employees, in October, 1998. The company manufactures saw blades. The blades are machined with cutting oils and the oil must be removed before the diamond paste is added to the teeth of the saw. If any contaminants remain, they can react in the furnace where the diamond powder is cured and ruin the blade. Barranca had used a PERC degreaser to clean the parts.

The SCAQMD had just added PERC to their toxics rule, Rule 1401. The permit on the company's PERC degreaser had lapsed and a new permit would be needed before the degreaser could be used. Mr. Delahaut was told by SCAQMD that he could not obtain a permit on the very old existing vapor degreaser because emissions of PERC from this degreaser would pose an unacceptable risk to the surrounding community.

Barranca Diamond Products could have purchased a new vapor degreaser to ensure that the operation posed a lower risk to the community. The cost of an airless/airtight degreaser for this operation would have been at least $200,000. Even if Barranca had the funds to make this purchase, there would have been other disadvantages to using PERC. The workers would be exposed to the chemical, a spill could contaminate the site and the company could have been sued under the Proposition 65 bounty hunter provision. At IRTA's recommendation, Mr. Delahaut made the decision to test water-based cleaners rather than continue to use PERC. IRTA began working with the company to assist in identifying, testing and implementing a water-based cleaning process.

Through laboratory testing, IRTA identified a few suitable cleaners that would be able to remove the contaminants from Barranca's saw blades. One of these cleaners was tested at the facility for a period to determine the parameters of the cleaning process. IRTA recommended that the facility test a spray cabinet which was likely to be the lowest cost equipment option that would be effective. The testing of the cleaning agent in the spray cabinet at the facility revealed that the process could be one step. That is, only a wash step was necessary; no rinsing or drying was necessary to produce the cleanliness required for the subsequent coating operation.

Barranca purchased the spray cabinet and has been using the water-based cleaning agent for several months. IRTA took a sample of Barranca's spent bath when it needed to be changed out after two months of operation. The sample met the discharge requirement limits. The Los Angeles County Sanitation Districts granted permission for the facility to discharge the cleaner periodically when it was spent as long as it met discharge standards. Barranca has no disposal costs for the water-based cleaner.

It is difficult to perform a comparative cost analysis of this case study. Under the new owner, Barranca never used the PERC degreaser so the costs of this operation are unknown. In addition, because the permit had lapsed, a new permit would have been required to use the solvent. Because of the new regulatory requirements, in order to use PERC, the firm would have had to purchase an expensive degreaser. The costs that are compared below are the actual costs of the water-based cleaning system and the hypothetical costs that the firm would have incurred had they decided to continue using PERC.

The cost of the 100-gallon capacity spray cabinet was $9,000. Figure 3-11 shows a view of the equipment. It is sized to hold the largest saw blades processed by Barranca which are 36 inches in diameter. Mr. Delahaut paid cash for the cleaning unit. Assuming a 10-year equipment life, the cost of the spray cabinet amounts to $900 per year.

The minimum price of a vapor degreaser that would meet the NESHAP and SCAQMD Rule 1401 requirements and have a capacity of 27 cubic feet is about $200,000. It holds 200 gallons of PERC. If Mr. Delahaut had purchased the vapor degreaser, he would have obtained a loan. Assuming a six percent cost of capital and a 10-year equipment lifetime, the annualized cost of the vapor degreaser purchase is $27,174.

Barranca uses a water-based cleaner, Daraclean 200, which is certified as a Clean Air Solvent. The cost of the water-based cleaning concentrate is $10.90 per gallon. Barranca uses it at a seven percent concentration. Assuming that the bath requires changeout every 45 days, then the annual cost of the cleaning agent for the eight changeouts is $610. A small amount of additional cleaner is required to replace evaporation. Assuming this is 10 percent of the amount required for changeout, the total annual cost of the cleaning agent is $671.

The cost of PERC is currently about $6.75 per gallon. Airless/airtight degreasers are very conservative of solvent. If the firm purchased such a degreaser, the emissions of PERC

would amount to only about 26 gallons annually. The spent PERC bath might require changeout twice a year. Replacement PERC for the emissions and changeout amounts to 426 gallons annually. The cost for replacement PERC is $2,876. The PERC is classified as hazardous waste and the disposal cost is $100 per drum. Assuming the PERC is changed out twice per year, the annual disposal cost amounts to $800.

The electrical requirements are higher for the water-based cleaning system than for the PERC degreaser. Barranca cannot accurately separate the electrical cost of running the spray cabinet from the cost of running the facility's other equipment. The electrical cost for running the spray cabinet is estimated at $80 per month or $960 per year. Although the vapor degreaser would use energy, for purposes of analysis, it will be ignored here.

The Barranca cost comparison is summarized in Table 3-7. Additional costs for the vapor degreaser that are not included in the table are the annual cost of the air district permit for the vapor degreaser and the fees for the PERC emissions. Including these costs would make the total cost for the PERC vapor degreaser even higher.

The values show that the cost of using the PERC vapor degreaser is about $30,000 per year. The cost of using the water-based cleaning system is about $2,500, or one-twelfth the cost. Other benefits with the water-based system are that there is no worker or community exposure to a suspect carcinogen and the firm is not vulnerable to bounty hunter suits and negative publicity.

A stand-alone case study for Barranca is included in Appendix B.

3.2.2. Electronics Cleaning

Printed circuit board assembly processes involve soldering components to the bare boards. A flux is applied to the boards to effect heat transfer in the soldering process and to facilitate solder flow. The components, like resistors and capacitors, are soldered either in holes that have been drilled in the board or to the surface of the board. Many companies make so-called through-hole boards and many make surface mounted boards. Most boards have combinations of through-hole and surface mount components.

After the components have been soldered to the boards, the flux and any other contaminants that might be present must be removed from the board. Traditionally, TCA or CFC-113, generally in combination with small amounts of alcohol, were used in printed circuit board defluxing. Prior to the production ban, companies began investigating alternative methods.

There are three types of flux that can be used on the boards. First, a very low solids (3 to 5 percent) flux, can be used. These fluxes do not have to be cleaned because they leave very low residues on the boards. Assemblers that make boards for the commercial market often use low solids flux and many assemblers for the commercial market converted to these fluxes when it was clear that production of TCA and CFC-113 was to be banned.

The second type of flux that can be used is organic acid or water soluble flux. This type of flux can be removed with plain D.I. water. The third type of flux that can be used is rosin flux. Rosin flux must be removed with either a solvent or a water-based saponifier. Until a few years ago, all military contracts required the use of rosin flux. At this stage, new military contracts allow the use of water soluble flux but some existing contracts still require the use of rosin flux. Most military contractors still will not allow the use of low solids flux.

The advantage of using low solids flux is that it need not be cleaned. Virtually all commercial operations could, in principle, convert to this type of flux. The advantage of water soluble flux is that it can be cleaned with D.I. water; the systems can be completely close looped and the cleaning operation is much lower cost than the rosin flux system. With rosin flux, only the rinse step can be close looped and often the rosin flux saponifiers contain VOC solvents. In recent years, a few rosin flux saponifiers without solvent additives have been developed.

Two of the companies featured here converted away from vapor degreasing solvents used for printed circuit board defluxing. Hydro-Aire, an aerospace subcontractor, was using TCA for cleaning boards; the company converted to a water-based defluxing process. Aerojet Electronic Systems, a large aerospace firm, converted from a CFC-113 defluxing process to a water-based cleaning system. Both companies avoided converting to VOC solvents for their cleaning needs by using water-based saponifiers with no solvent additives.

3.2.2.1. Hydro-Aire

Hydro-Aire was using TCA for removing the flux from printed circuit boards after soldering components to them. This operation had a military specification that required the use of rosin flux. A vapor degreaser was used to remove the flux from the boards after soldering. Hydro-Aire had another commercial printed circuit board operation that used water soluble flux. In this operation, the company used plain D.I. water to clean the flux from the boards in a small dish washer.

In 1998, Hydro-Aire purchased a large batch dishwasher from Aqueous Technologies and adopted a water-based cleaner made by Church & Dwight. This cleaner, unlike many others used for removing rosin flux, contains no solvent additives. A picture of the new system is shown in Figure 3-12. The cost of the new cleaning system, including racks and tax, was $24,950. This cleaning system is now used for cleaning both the commercial and the military printed circuit boards. Hydro-Aire also purchased an evaporator, at a cost of $6,950, to use for disposal of the water-based cleaners. The total cost of the cleaning unit and evaporator including associated equipment and taxes was $37,378. Hydro-Aire paid cash for the new system. Assuming a 10-year useful life of equipment, the annual cost of the new system is $3,738.

In the year prior to the conversion, the cost of the TCA purchased for the printed circuit board operation was $25,000. The cost of the deionized water used in the second commercial machine was $4.

The cleaning system is operated for an average of five cycles per day. The length of a cycle is about 50 minutes. The machine takes about 10 minutes to heat the 5-gallon tank of formulation. It requires 5 minutes for the wash step; it goes through four 5-minute rinse steps and a 15-minute drying step. The water-based cleaner is used in a 10 percent concentration in the 5-gallon wash bath and each time the unit is operated, a new cleaning bath is required. Assuming 260 days per year of operation, the amount of cleaner required is 650 gallons per year. At a cost of about $12 per gallon, the annual cost for the water-based cleaning agent is $7,800. Deionized water is required for rinsing the boards; Hydro-Aire estimates this cost at about $20 per year.

Disposal costs for the spent TCA from the cleaning operation are estimated at $400 annually. The sludge from the water-based cleaning formulation that remains after evaporation requires disposal. Assuming a 5 percent contamination level and the evaporation of 25 gallons per day, the amount of contaminants that require disposal are 325 gallons annually. This sludge contains lead and is classified as hazardous waste. At a cost for disposal of about $200 per drum, the total annual cost of the sludge disposal is $1,200.

About half the 14,000 boards were cleaned in the TCA system and half in the deionized water system before the conversion. Both the new and old equipment was automated. The workers required 4 minutes labor time to clean 7,000 boards in the TCA degreaser. They required 1 minute labor time to clean the remaining 7,000 boards in the deionized water system. With the change to the new system, the labor required was 1 minute for all 14,000 boards. At a labor rate of $19.50 per hour, the total annual labor cost before the conversion was $11,375. The total annual labor cost after the conversion was $4,550.

The electricity cost for the vapor degreaser and the deionized water unit are not known and will be assumed to be negligible for purposes of analysis. On the new system, the pump power is 600 watts. The pump runs during for the wash and rinse steps for 25 minutes per cycle. The total annual electrical use for the pump is 325 kW. The power for the blower in the dryer is 10.4 kW. The dryer operates for 15 minutes each cycle. The annual electrical use from the dryer is 3,380 kW. The evaporator is assumed to operate 8 hours per day. The heater power in the evaporator is 11 kW. The annual electrical use from the evaporator amounts to 22,880 kW. Assuming a cost of 12 cents per kW, the total annual electrical cost for operating the new system is $3,190.

Regulatory fees for the TCA include emission fees and permit fees. About 200 gallons of TCA or 2,200 pounds were emitted annually. At a cost of 4 cents per pound, the annual cost for emissions of the TCA amounts to $88. The annual permit renewal fee for the degreaser was $179. The total regulatory costs are $267. This does not include any costs that might be incurred for recordkeeping and reporting.

Table 3-8 presents the cost comparison for the TCA degreaser and the new water-based cleaning system. The conversion reduced the costs for cleaning printed circuit boards by 45 percent.

Hydro-Aire reduced their emissions of TCA by 2,200 pounds or 1.1 tons per year. The savings to the company for the conversion amounted to $15,044 per ton of TCA reduced.

A stand-alone case study for Hydro-Aire's printed circuit board cleaning conversion is presented in Appendix B.

3.2.2.2. Aerojet Electronic Systems

Aerojet is an operating segment of GenCorp. The company designs and manufactures space surveillance, meteorological sensor and smart weapon systems at its Azusa plant. As part of their operation, they assemble printed circuit boards that are used in the equipment the company manufactures.

About five years ago, Aerojet completed their evaluation of alternatives to TCA and CFC-113. The company had been using various CFC-113 blends in a vapor degreasing process for removing flux from printed circuit boards. Aerojet conducted an extensive set of tests to determine which technology they should adopt. Since the company had a number of military contracts, they were required to use rosin flux.

Aerojet determined early on that they would focus the research on water-based cleaning alternatives. The company tested water soluble flux and found it could be effectively removed with D.I water. The company also identified a rosin flux saponifier based on a Reactive Aqueous Defluxing System. The cleaner, called RADS, in contrast to other saponifying cleaners available at the time, did not contain solvent additives.

Aerojet decided to use the RADS technology with rosin flux. They thoroughly investigated cleaning and waste water treatment equipment that would be flexible enough so the process could one day be converted to water soluble flux. That process is attractive because the boards can be cleaned with plain D.I. water and the entire system can be close looped.

After evaluating a range of equipment options, Aerojet purchased a conveyorized custom designed system that could be adapted to water soluble flux removal. Although the equipment has other stages, the stages that are being used today for cleaning include a wash bath which contains the RADS cleaner, a rinse bath that contains D.I. water and a final rinse bath that contains D.I. water. The final rinse recirculates to the first rinse so the boards see the purest water at the end of the cycle.

The company was committed to zero discharge and investigated methods of recycling and disposal. In the end, Aerojet purchased a closed loop recycling system that utilizes a mixed ion bed filtration technology; this system is used to process the rinse water. It cleans the water and recirculates it back to the rinse chamber. Aerojet also purchased an evaporator that treats the spent wash water when it can no longer be used.

Aerojet intends to examine water soluble flux in the future. At this stage only one of their existing military contracts still requires the use of rosin flux. The Aerojet system can easily be adapted to clean water soluble flux.

3.2.3. Precision Cleaning

One of the companies that is featured here, Astro Pak, has converted from a vapor degreasing process using NPB to a water-based cleaning process for precision cleaning. Another company, Kaiser Electroprecision, is still in the process of testing a water-based cleaner that would replace TCA used in a precision cleaning process. Precision cleaning differs from other cleaning operations because a cleanliness standard is usually specified. The standards are generally of two types: nonvolatile residue analysis (NVR) and particulate count.

3.2.3.1. Astro Pak

Astro Pak is one of the nation's leading precision cleaning contractors specializing in the cleaning of high purity gas and fluid systems. The company provides precision cleaning services and cleanliness certification for pipes, valves, tubing, components, tanks, hoses and fittings for almost every industry, including aerospace, military, pharmaceutical, microelectronics and semiconductor.

There is a perception that solvent cleaning agents are required for precision cleaning operations. The reasoning is that water-based cleaners are appropriate for general metal cleaning but cannot clean as effectively as solvents for precision applications. The Astro Pak case study described here demonstrates that water-based cleaners perform as well as or better than solvents in this Astro Pak's precision cleaning application.

Astro Pak used TCA for many years in a vapor degreasing process. The company converted to NPB, which is classified as a VOC, in late 1997. They used the NPB for about a year. As part of this project, IRTA assisted Astro Pak in converting to a water-based cleaning process.

Astro Pak purchased an ultrasonic water-based cleaning system that consisted of an ultrasonic wash bath and two immersion rinse baths. Each of the baths hold about 700 gallons. The cost of the system, including installation, was $171,788. The company amortized the cost over a four year period. Assuming an interest rate of 7.75%, the annual cost for the equipment is $50,085.

Astro Pak's annual purchase cost for NPB was $66,240. At a cost of $2,550 per drum, Astro Pak used about two drums of the chemical per month. Astro Pak is using a water-based cleaner, Daraclean 212, which has aerospace approval and is certified as a Clean Air Solvent. The cost of the aqueous cleaning formulation is $2,160 annually. This assumes that the bath must be changed out every four months.

The cost of gas that was used with the NPB vapor degreaser was $3,297 per year. The increase in the cost of the electricity after the water-based cleaning system was installed was $684 per month or $8,208 annually. The cost of supplying deionized water for the new water-based system is $225 per month or $2,700 per year.

The SCAQMD regulatory fees for the vapor degreaser permit renewal and the VOC emissions amounted to $631 annually.

Table 3-9 summarizes and compares the annual cost of using the NPB vapor degreaser and the water-based cleaning system. Conversion to the water-based cleaning system reduced the costs by 10 percent. Moreover, after the water-based cleaning equipment is paid off in four years, the annual costs for using the water-based cleaning system declines to $13,068 which is less than one-fifth the cost of using the vapor degreasing system.

When the water-based cleaning system was installed, Astro Pak began conducting a study to compare the cleaning capability of the new water-based cleaner with the NPB vapor degreasing process. For this comparison, Astro Pak measured the nonvolatile residue (NVR) remaining on different types of hardware after they were cleaned with the new water cleaner. Astro Pak selected hardware for which they had already measured the NVR after cleaning with NPB. NVR is a cleanliness verification test and is a measure of the nonvolatile residue remaining on the part after cleaning. The higher the NVR, the more contaminated the part.

Astro Pak measured the NVR on 25 different pieces of hardware cleaned with the water system and NPB. Of the 25 pieces tested, 21 had a lower NVR when they were cleaned with the water-based system. Three of the pieces had the same NVR. Only one piece had a lower NVR when it was cleaned with the NPB. The average NVR level achieved when the hardware was cleaned with NPB was 0.756 milligrams per square foot. The average NVR level achieved with the water-based cleaning system was 0.616 milligrams per square foot. Astro Pak concluded that the water-based cleaner is as effective or more effective than the NPB.

A paper scheduled to be published in "Precision Cleaning Magazine" in October provides more detail on the Astro Pak conversion. A stand-alone case study is also included in Appendix B.

3.2.3.2. Kaiser Electroprecision

Kaiser Electroprecision is an aerospace electronics company located in Irvine. The company makes high pressure, breathing oxygen valves that are used in liquid oxygen systems in Boeing aircraft and in ground support equipment.

The oxygen valves are made of brass, aluminum or stainless steel. The valve components are lapped and are assembled in a cleanroom environment. As part of the assembly process, the valves are cleaned in cold TCA in an ultrasonic cleaning system to remove the lapping compound. They are dipped in a phosphoric acid etch solution and dried. The valves are then taken into a cleanroom and where they are cleaned again with TCA and dried with nitrogen. This final cleaning process is used to remove fingerprints and particles.

After the process, the valves must meet stringent particle count and NVR standards. Kaiser Electroprecision sends the parts off-site for NVR analysis. The NVR standard for acceptable valves is 3 milligrams of hydrocarbons per square foot.

Kaiser Electroprecision wanted to investigate alternatives to TCA and IRTA began work with the company. Production of TCA has been banned because the chemical contributes to ozone depletion. Eventually the supplies of TCA will be depleted and Kaiser Electroprecision will no longer be able to purchase it. The company did not want to convert to a VOC solvent for the cleaning process.

IRTA arranged for laboratory testing of the parts and two water-based cleaners with aerospace approval that could be suitable for Kaiser Electroprecision's cleaning application were identified. One of these, Daraclean 212, has been certified as a Clean Air Solvent by SCAQMD. It was selected for further on-site testing. This is the same cleaning agent Astro Pak converted to for precision cleaning.

Because Kaiser Electroprecision's valves had blind holes, it was apparent that the company would require an ultrasonic cleaning unit. IRTA arranged for a vendor to provide the company with a small heated ultrasonic unit for testing. The Daraclean 212 was tested at the facility and the parts were evaluated for cleanliness. As mentioned above, the NVR cleanliness standard for the valves is 3 milligrams per square foot. The valves that were cleaned in the cleanroom at Kaiser Electroprecision with the water-based cleaner failed the NVR test.

As noted in the last section, Astro Pak uses Daraclean 212 and is able to clean their hardware effectively enough to achieve NVR levels well below 1 milligram per square foot. This implies that Kaiser Electroprecision should be able to do the same. After discussion with Astro Pak personnel, it was concluded that the problem was likely in the quality of the deionized water Kaiser Electroprecision used for rinsing the parts during the testing. The D.I. water is transported in pipes from another location in the plant and it may not be as pure as necessary. At this stage, Kaiser Electroprecision is testing their D.I. water quality. If it is found to be inadequate, the parts will be cleaned with the Daraclean 212 again, rinsed with the higher quality D.I. that is required and tested for NVR.

The water-based process should be suitable for Kaiser Electroprecision because it is already working well for Astro Pak. Astro Pak's hardware has NVR levels well below the 3 milligram per square foot required for Kaiser Electroprecision's parts. Rinsing with high quality D.I. water is obviously an important part of the process.

 SCAQMD Rule 1124 regulates VOC emissions from aircraft and spacecraft coating, assembly and cleaning operations. The rule specifies that the solvents used for cleaning or cleanup must have a VOC composite partial pressure of 45 mm Hg or less at 20 degrees C or the solvent must contain 200 grams per liter VOC or less. Materials used for stripping aerospace components must contain less than 300 grams per liter VOC or have a VOC composite partial pressure of 9.5 mm Hg or less at 20 degrees C.

When the District was investigating a VOC RECLAIM program, an inventory of aerospace VOC emissions was developed. The total VOC emissions from coatings were estimated at 737 tons per year; VOC emissions from cleanup solvents were estimated at 442 tons per year. At the time, TCA was widely used in both coatings and cleanup solvents. The District estimated the use of ozone depleting compounds, which may also have included some CFC-113, at 2,400 tons per year. There are approximately 250 aerospace companies in the South Coast Basin and only 74 of them were included in VOC RECLAIM development. Thus the estimates of VOC and ozone depleting handwipe emissions were an underestimate of the actual emissions from all aerospace facilities.

Since that time, aerospace companies have converted away from TCA and CFC-113 in coatings and cleanup solvents to other products which are generally VOC containing materials. During RECLAIM development, one representative from an aerospace company stated that the total VOC handwipe emissions estimated by the District were equal to the emissions from his facility alone. IRTA talked to several other aerospace company representatives at that time and determined that handwipe emissions in aerospace facilities represent between 50 and 80 percent of total facility emissions.

If the VOC RECLAIM data are correct and firms simply substituted VOC solvents for ozone depleters, the 74 RECLAIM facilities would emit 2,842 tons of VOC solvents from handwipe operations each year. If some companies minimized their usage of handwipe solvents and converted to exempt or non-VOC alternatives, then the number could be much lower. On the other hand, there are far more than 74 aerospace companies in the Basin which would suggest that the emissions could be even higher than 2,842 tons per year. For purposes of analysis in this report, the figure of 2,842 tons of VOC per year or 11.4 tons per day is used.

Aerospace handwipe operations are widely varied. They range from wiping down aircraft skins prior to coating to final assembly cleaning for satellites. Conversion to alternatives must generally be accomplished on a case-by-case basis. Water-based cleaners are appropriate for some applications and acetone is effective for some others.

IRTA worked with one aerospace company, Hydro-Aire, on handwipe cleaning and a small stripping operation. The analysis of these examples is included here. Also included in this section is information on a handwipe conversion made by the Rocketdyne division of Boeing a number of years ago. Information on these two companies is presented below.

4.1. Hydro-Aire

Hydro-Aire had several operations where solvents are used for handwipe. The first operation involves handwiping in the laboratory area. The company used methyl ethyl ketone (MEK) for cleaning ink and penetrant from parts that had gone through penetrant inspection for leaks. IRTA suggested the facility test acetone. The acetone proved to be a good alternative and Hydro-Aire made a conversion. About 2 gallons of MEK were used in the handwipe operation each month. The amount of acetone used in the handwipe operation is also 2 gallons per month. Hydro-Aire pays $5.40 per gallon for MEK and $4.33 per gallon for acetone. The company realized a cost savings of about $26 annually through the conversion.

Hydro-Aire used a small amount of methylene chloride to clean cured coatings from coupons. IRTA suggested the company try formic acid as an alternative. The tests of the formic acid were successful. Hydro-Aire was unable to provide information on the amount of methylene chloride that was used for this purpose. IRTA estimates the use of methylene chloride at 5 gallons a month. Hydro-Aire pays $36 per gallon for methylene chloride. The price of the formic acid used in the testing is $39 for a 5-gallon container or $7.80 per gallon. It is estimated that the formic acid use would be only three-fourths that of methylene chloride, or 3.75 gallons per month, because formic acid is much less volatile than methylene chloride. Thus the annual cost for purchasing the methylene chloride is $2,160 and the cost for purchasing the formic acid is $351. A conversion to formic acid, a non-VOC, would reduce toxic emissions by 660 pounds or 0.33 tons per year and it would reduce Hydro-Aire's costs.

Hydro-Aire used a VOC solvent blend of glycol ethers to activate the plastic in a wire bonding application. IRTA suggested the company try acetone and the tests were successful. The cost of the VOC solvent blend was very high, at $59 for about one pint. One of these containers was used about every two weeks. The total annual cost of purchasing the VOC solvent was $1,534. The cost of acetone is about $2.10 per gallon. Assuming the same amount of acetone would be required, the total annual cost for using acetone in this operation is about $7.

4.2. Rocketdyne Division of Boeing

Rocketdyne manufactures the space shuttle main engine at their location in Canoga Park. The proper operation of this engine is dependent upon specified cleanliness levels which are established and verified during final precision cleaning operations. These requirements are important in the assembly, testing and handling because contamination could lead to blockage of critical orifices in the propellant flow streams and uncontrolled ignition reactions with the oxidizer propellant. Either of these events could cause engine failure.

For many years, Rocketdyne used CFC-113 and TCA for handwipe cleaning during various assembly, test and handling operations and also for final precision cleaning and verification of cleanliness. Production of these two cleaning agents was banned in January, 1996. Rocketdyne established programs to evaluate alternatives in handwipe cleaning and for verification of cleanliness and these two evaluation programs are discussed below. As mentioned earlier, the current limit in Rule 1124 for cleaning solvents is a VOC content of 200 grams per liter or 45 mm Hg composite vapor pressure. Rocketdyne's operations are not subject to these limits because there is an exemption in the rule for cleaning space vehicle hardware.

4.2.1. Handwipe Cleaning

Rocketdyne identified over 200 aqueous and solvent candidates as replacements for CFC-113 and TCA for cleaning the engines. Some of these cleaning agents were also tested for final cleanliness maintenance handwipe operations. About one-fourth of these, or 55 cleaning agents, were tested. The evaluation had three phases. The first phase involved examining:
o the environmental, health and safety of the alternative
o materials compatibility with nickel and titanium, two materials that are used extensively by Rocketdyne
o the cleaning effectiveness of the cleaner

Of the 55 cleaners evaluated, 52 were not taken beyond the Phase I testing. Thirty-eight failed because they had unacceptable information listed on the MSDS or technical literature related to environmental, health or safety concerns, material incompatibility or unacceptable chemical or physical property data. Eleven failed the compatibility tests with nickel and/or titanium. Two failed because they were not effective in cleaning Rocketdyne's contaminants. The manufacturer stopped producing one of the cleaning agents which eliminated it from further consideration.

The three cleaning agents remaining, which included two aqueous cleaners and Vertrel XF or HFC-43-10, a hydrofluorocarbon made by DuPont, were put through the Phase II testing phase. Phase II testing involved :
o measurement of the nonvolatile residue (NVR) left after cleaning
o compatibility testing with non-metallic materials and metallic alloys including oxygen which is the oxidizer propellant used for powering the engine
o a series of tests to confirm chemical and physical properties

The two aqueous cleaners failed the Phase II testing because of the NVR tests results. These cleaners contain surfactants which leave an unacceptable residue on the part.

The Phase III testing was a user evaluation. The HFC-43-10 passed this testing phase and Rocketdyne converted to the solvent for handwipe operations. The solvent does not contribute to ozone depletion and it showed similar or better cleaning capability than the handwipe operations used by Rocketdyne. HFC-43-10 has a high vapor pressure so it evaporates rapidly, has no flash point and is compatible with all the materials used in the engine including oxygen at ambient and high pressures.

4.2.2. Cleanliness Verification

TCA was used by Rocketdyne for cleanliness verification of the engine hardware. Cleanliness verification is a quality assurance procedure. After flushing critical hardware surfaces, an amount of the final flush solvent is collected, filtered for particulate analysis and gravimetrically analyzed for NVR. Rocketdyne needed an alternative to TCA for the verification process.

Rocketdyne first tested an aqueous method developed by NASA and adapted it for their operations. It utilizes a total organic carbon analyzer (TOCA). The hardware is cleaned in a heated ultrasonic system containing deionized water. A portion of the water is injected into the TOCA to determine the carbon content. A low carbon content signifies a low NVR. In laboratory testing, this method held promise and Rocketdyne conducted additional testing on the hardware to demonstrate that the technique was feasible and to qualify it for production use.

Rocketdyne also investigated an aqueous method for particulate verification. This method involved collecting 500 milliliters of the final deionized water used to rinse the part. The water is filtered and the filter media is analyzed for particulate matter.

Solvent methods for the NVR analysis were evaluated for use where the aqueous method was not appropriate. This includes cases where the hardware is large or the items are damaged by the ultrasonic action. Rocketdyne evaluated a number of substances and tested their effectiveness in removing contaminants. The three top performing solvents were cyclohexane, ethyl acetate and HCFC-225. The HCFC is exempt from VOC regulation and cyclohexane and ethyl acetate are classified as VOCs. HCFC-225 has a worker exposure level of 50 ppm and it was eliminated from further consideration because of its toxicity and because it contributes to ozone depletion and will eventually be banned for this reason. Ethyl acetate was eliminated from further consideration because it did not have high solubility for waxes and greases. Cyclohexane showed cleaning efficiencies similar to TCA for the contaminants and it was selected as the candidate to replace TCA.

Rocketdyne performed extensive testing of the aqueous NVR method, the aqueous particulate method and the cyclohexane NVR method on their hardware. They contaminated hardware with typical contaminants and compared the new materials with TCA for the verification testing. For single contaminant NVR testing, baseline TCA was 82 percent effective in detecting the contaminants. The aqueous method and the cyclohexane method were 83 percent effective in detecting the contaminants. For a contaminant mixture, TCA was 88 or 89 percent effective in detecting the contaminants. The aqueous method was 91 percent effective and the cyclohexane was 89 percent effective.

The hardware testing for the aqueous particulate method indicated that the aqueous method was as efficient as TCA in removing particulates in the size range of interest to Rocketdyne. In only one case the TCA removed a larger number of particles but this was not significant. Use of the aqueous particulate method allowed Rocketdyne to eliminate an intermediate drying step since it is performed at the end of the cleaning process.

The new verification methods have some limitations. Only hardware with dimensions smaller than about 12 inches by 12 inches can be used with the aqueous NVR verification method. In addition, hardware with dry film lubricants cannot be subjected to the ultrasonic agitation. It is acceptable to use the method on anodized or chem filmed aluminum for durations of 10 minutes or less. The aqueous method cannot detect silicone greases, fluorinated greases or paraffin waxes. It should be noted that the TCA method cannot detect the silicone or fluorinated greases either.

The cyclohexane methods should not be used on hardware containing neoprene, silicone, butyl or ethylene propylene elastomers because these materials are degraded by the solvent. Cyclohexane, like TCA, is not effective in detecting silicone or fluorinated greases. The two main drawbacks of cyclohexane are that it has a flash point and it is a VOC. Rocketdyne designed a special glove box system that employs gaseous nitrogen as a purge gas for using the cyclohexane. An oxygen analyzer monitors the oxygen content inside the system. Cyclohexane is not compatible with oxidizers and a vacuum oven drying method is used on the hardware that is flushed with the solvent.

A handwipe cleanliness verification method using cyclohexane is used for the hardware that cannot be verified by the aqueous and glove box method because they are too large or because of material incompatibility. Rocketdyne tested dry wipes, IPA, D.I. water and cyclohexane and cyclohexane was selected for this application.

V. RULE 442--USAGE OF SOLVENTS

Operations involving solvent usage that do not fall under any of the District's source specific coating or cleaning rules or are specifically exempted are covered by SCAQMD Rule 442. According to this rule, photochemically reactive solvent used for industrial and commercial surface cleaning or degreasing operations must be controlled. The emissions must be reduced by at least 85 percent by weight. The rule also requires that organic materials emitted to the atmosphere from the use of photochemically reactive solvents are limited to 7.9 pounds per hour, not to exceed 39.6 pounds per day. Organic materials emitted into the atmosphere from the use of non-photochemically reactive solvents are limited to 81 pounds per hour, not to exceed 600 pounds per day.

The District defines "photochemically reactive solvent" in SCAQMD Rule 102. It means any solvent with an aggregate of more than 20 percent of its total volume composed of the chemical compounds classified below or which exceeds any of the following percentage composition limitations:
o a combination of hydrocarbons, alcohols, aldehydes, ethers, esters or ketones having an olefinic or cycloolefinic type of unsaturation except perchloroethylene: 5 percent.
o a combination of aromatic compounds with eight or more carbon atoms to the molecule except ethylbenzene, methyl benzoate and phenyl acetate: 8 percent.
o a combination of ethylbenzene, ketones having branched hydrocarbon structures, trichloroethylene or toluene: 20 percent.

During this project, IRTA worked with one company, Nelson Name Plate, that had an operation that was not covered by any of the District's Regulation XI rules. This operation is discussed in more detail below.

5.1. Nelson Name Plate

Nelson manufactures name plates from aluminum, stainless steel and brass stock. The name plates go through several operations as part of the manufacturing operation. Nelson has a cleaning operation for removing the oil from the stock, a lithographic printing operation, a screen printing operation and a photoresist/coating stripping operation. In this section, Nelson's stripping operation is discussed. The photoresist is applied to part of the plate. The plate is then painted with one of several coatings used by the company.

For several years, Nelson used TCA as a stripping agent to remove the coating and the photoresist from part of the plate. Some 500 gallons per month or 6,000 gallons per year were used in the photoresist stripping operation. At a cost of $16.15 per gallon, the cost of purchasing the TCA was $96,900 annually. TCA is exempt from VOC regulations and its use is not covered by Rule 442.

Because of the production ban on TCA, IRTA began work with Nelson to identify a suitable alternative. Acetone blends were tested and the 1.60 pound per gallon blend used in lithographic press cleanup described in Section II proved suitable for this operation. IRTA began testing water-based alternatives to reduce the VOC content further and because Nelson would rather use water-based than solvent products where possible. While the work on water-based cleaners continued, Nelson converted to the 1.60 pound per gallon acetone blend.

The cost of the 1.60 pound per gallon blend is $3.70 per gallon. Usage of the stripper for this process was reduced to 320 gallons per month or 3,840 gallons per year. The annual cost for purchasing the new stripper is $14,208.

Nelson generates a small amount of waste sludge from the process. The amount of sludge generated and the disposal cost did not change with the conversion from TCA to the low VOC content blend.

Nelson paid emission fees to the SCAQMD for TCA. At a fee rate of three cents per pound, the cost of emitting TCA amounted to $1,980 annually. The cost of emitting the VOC solvent blend, assuming a VOC emissions fee of $292.80 per ton, amounts to $882 per year.

Table 5-1 shows the cost comparison for the TCA and low-VOC content stripper blend. Nelson has reduced the cost of stripping by 85 percent through the conversion.

5.2. Other Rule 442 Facilities

IRTA is aware of two companies with very high VOC emissions that fall under Rule 442. One of these companies, Stewart Film Screen, manufactures movie screens using large amounts of VOC containing coatings. Emissions from this operation are estimated at about 1 ton per day. A second company, MCP Foods, uses isopropyl alcohol (IPA) as an extraction/evaporation solvent in a food processing operation. Emissions from this operation are also estimated at 1 ton per day.

There are likely to be many more operations in the Basin that are regulated by Rule 442. This rule allows very high VOC emissions and the emissions from Rule 442 operations could be very large.

During this project, IRTA identified four other sources of VOC emissions that may be very large. These include the use of isopropyl alcohol (IPA) for cleanroom maintenance, the use of janitorial cleaning products which contain VOCs of various types, the use of mineral spirits or kerosene to dilute machine oils and the use of vanishing oils. The emissions from these sources are not known but they are likely to be very high. Each of these categories is discussed below. This section also identifies another area where VOC emissions could be reduced.

6.1. Cleanroom Maintenance

Hundreds, perhaps thousands, of companies in the South Coast Basin have one or several cleanrooms. Cleanrooms are designed to minimize or eliminate particulate contamination from the manufacture and assembly process. These contaminants could cause sensitive components to fail at a later date while they are in operation. The types of companies that have cleanrooms include semiconductor manufacturers, disk drive manufacturers, electronic component manufacturers, aerospace firms and medical device manufacturers.

Various levels of cleanliness are required in the manufacturing process, depending on the type of product that is manufactured. These cleanliness levels are classified in terms of the number of particles present in a volume of air. For example, in a Class 100,000 cleanroom, there must be less than 100,000 particles measuring 0.5 micron or larger in a cubic foot of air. Semiconductor facilities often have Class 1 cleanrooms which require the highest level of particulate control. Aerospace facilities may have 1,000, 10,000 or 100,000 cleanrooms.

Companies that have cleanrooms maintain them regularly by wiping down all the surfaces of the cleanroom including the floors, ceilings, walls and workspace surfaces daily. When a cleanroom is being established, it may require several days of wipedown to achieve the low particulate levels that are required. Many companies use IPA or blends of IPA with deionized water to establish and maintain their cleanrooms. Some companies use plain deionized water for their wipedowns.

In 1996, IRTA began working with a company, Spaceport Systems International (SSI), that wanted to establish and operate a cleanroom at Vandenberg Airforce Base. SSI was to provide satellite booster processing and commercial satellite launches. The company was informed that there would be a VOC cap on their emissions. IRTA initiated a test and demonstration project to identify non-VOC cleanroom solvents that would be effective in particulate control.

A paper describing the procedure for the testing and the results of the tests is provided in Appendix C. The cleaning agents that were tested had to be:
o non-VOCs
o non-ozone depleting
o low in global warming potential
o not on toxics lists
o relatively low in toxicity
o commercially available

After screening cleaning agents according to these criteria, the alternatives that were to be tested were selected. They included deionized water (D.I.), various water-based cleaners, acetone and HFC-43-10. IPA, in combination with D.I. in various blends was to be tested as the control.

The procedure involved preparing 22 stainless steel panels and allowing them to accumulate particles. These 22 panels were either not cleaned or cleaned with one of the chemicals or blends. A representative of the Palt Company, a firm that performs and arranges cleanroom maintenance, cleaned the panels in a flow bench in a cleanroom at the Rocketdyne Division of Boeing. A representative of Dryden Engineering tested the bottom third of each panels for particles using their particle counter. The number of particles per square inch in the 0.3 to 10 micron range was recorded for each panel. The panels were bagged and sent to National Technical Systems for non-volatile residue (NVR) testing. The top third of each panel was tested for NVR to determine if the cleaning agent left a residue.

The results demonstrated that the low-IPA content formulation (10 percent IPA/90 percent D.I.) was more effective than the high IPA (70 to 90 percent IPA) blends at controlling particulates. The results also showed that plain D.I. water provided as effective particulate control as the best IPA blend. The water-based cleaners left a residue on the panels; they were judged to be poor cleaners because this residue, which could serve as a particle attractant, would have to be rinsed with plain D.I. water, adding another step. The acetone and HFC and blends of the two solvents performed relatively poorly. The high vapor pressure of the two solvents may have dried the wipes too fast to control the particles well. The paper in Appendix C presents the results in more detail.

Even though the test program conducted by IRTA demonstrated that plain D.I. water could control particulates as effectively as IPA, it is traditional to use IPA for that purpose. Some companies use only D.I. water for particulate control. Three of the companies IRTA worked with during this project, Litton Guidance & Control Systems, the Rocketdyne Division of Boeing and Astro Pak, use only D.I. water for cleanroom maintenance. One other company IRTA worked with, MiniMed, does use IPA for cleanroom maintenance.

Medical device manufacturers may require biocidal control for their products and their cleanrooms. IPA is the best method medical device manufacturers have found for biocidal control and the Food and Drug Administration requires its use for many applications. The industry has determined that biocidal control can be achieved with a 70 percent IPA/30 percent D.I. blend and this is commonly used. In some cases, medical device manufacturers use pure IPA or the blend described above for all their cleaning even where biocidal control may not be necessary.

MiniMed does not require biocidal control in their cleanrooms but the company uses pure IPA for cleanroom maintenance. A MiniMed representative estimates that the company uses 10 gallons of IPA per month at an estimated cost of $9 per gallon. The total cost of purchasing IPA amounts to $1,080 annually. Most medical device manufacturers have purchased mixed bed deionizing systems for generating D.I. water throughout their facility. Another option exercised by many companies is to rent deionizing systems from a service provider. Assuming a cost of $150 per month for this D.I. system, the cost of generating D.I. water would be $1,800 per year. The D.I. system would be capable of supplying D.I. for purposes other than cleanroom maintenance. If 40 percent of the D.I. water is used for other purposes, it is cost effective to adopt the plain D.I. water for cleanroom maintenance in place of IPA.

A representative of a company that supplies cleanroom maintenance equipment and supplies indicates that the alcohol used for cleanroom maintenance carries a price of $35 per gallon. His customers generally use a blend of 20 percent IPA/80 percent D.I. or 40 percent IPA/60 percent D.I. He also indicates that these companies use between 5 and 7 gallons of IPA per month. Assuming a company uses 5 gallons of IPA a month in a 20 percent blend and that the IPA cost is $35 per gallon, the cost for maintaining the cleanroom is $2,100 per year. The alternative would be to rent the D.I. system at $150 per month or $1,800 per year and eliminate the use of IPA.

Some companies that perform cleanroom maintenance in the South Coast Basin may emit more than 4 tons per year of VOCs from all of their facility operations. In this event, the IPA emitted from the cleanroom maintenance operation would be subject to emissions fees of $292.80 per ton. By converting to plain D.I. for this purpose, the company would also reduce their emission fees.

There is no information on the number of companies that perform cleanroom maintenance in the South Coast Basin and no information on what percent use IPA. IRTA gathered information on the number of facilities in the Basin that might have one or more cleanrooms and estimated emissions from this category. In the four-county area, there are 117 medical device manufacturers and perhaps three-fourths of them or 88 have cleanrooms. There are 250 aerospace companies in the Basin and again, it is estimated that three-fourths of them or 188 have cleanrooms. There are 1,018 electronics firms in the four-county area and it is estimated that one-third of them or 339 have cleanrooms. There are 9 disk drive manufactures and all of them likely have cleanrooms. There are 87 semiconductor manufacturers and perhaps half of them or 44 have cleanrooms. The total number of companies in the Basin that may have cleanrooms amounts to 730. Assuming that half of them use IPA for maintenance and that each uses 6 gallons of IPA per month, the total VOC emissions from cleanroom maintenance in the Basin is estimated at 173,448 pounds or 87 tons per year.

Many of the companies using IPA for cleanroom maintenance could convert to plain D.I. water. Some companies, medical device manufacturers who require biocidal control, would still have to use IPA for cleanroom maintenance. They could reduce their emissions, however, by converting to the 70 percent IPA/30 percent D.I. water blend that provides effective biocidal control.

6.2. Janitorial/Institutional Cleaning

Cleanroom maintenance cleaning agents may be part of a larger category of cleaners that could be classified as janitorial and institutional cleaning or as repair and maintenance cleaning. SCAQMD Rule 1171 defines institutional cleaning as "cleaning activities conducted at organizations, societies, or corporations including, but not limited to schools, hospitals, sanitariums, and prisons." Rule 1171 defines janitorial cleaning as "cleaning of building or facility components, such as the floors, ceilings, walls, windows, doors, stairs, bathrooms, etc." SCAQMD Rule 1171 provides an exemption for institutional and janitorial cleaning.

The California Air Resources Board (CARB) is responsible for regulating consumer products. The cleaners that are purchased for institutional and janitorial cleaning may not be consumer products. Rather, the companies that do the cleaning purchase the cleaning agents from manufacturers or distributors in large quantities. Whether these products are classified as consumer products is not clear. Even if CARB does have jurisdiction over this set of products, there is no reason the District could not begin regulating the cleaners in the South Coast Basin in cooperation with CARB.

Emissions of VOCs from institutional and janitorial cleaning are likely to be very high. Many companies add alcohols, glycol ethers or terpenes, which are all classified as VOCs, to their water-based products.

6.3. Machine Oil Dilution

Coolants are widely used by machine shops in the machines that are used to make metal parts. Petroleum based oils are often used by these machine shops. In some cases, the manufacturer sells the oil concentrate and recommends that the company dilute it with mineral spirits or kerosene prior to use. Other companies offer the oil preblended with the solvent. The petroleum based oils carry a cost of $3 per gallon for the concentrate.

Many companies, perhaps as many as 60 percent, have converted to synthetic or water-based coolants in their machines. These oils are water dilutable. The cost of these oils is about $5 per gallon and they are diluted in the 5 to 1 to the 30 to 1 range.

One company IRTA has worked with used petroleum based oil which was purchased as a concentrate. The company diluted the oil by using 10 parts kerosene to 1 part oil prior to using it in the machines.

Another company that participated in this project, S&H Machine Inc., uses petroleum based oil in the machines. This company converted their mineral spirits BLCCs to water-based cleaning processes as described in Section III. When the company used mineral spirits, the parts were blown off with shop air over the machines. The mineral spirits that ended up in the machine from this process were used to dilute the oil. Now that the company does not have this source of dilution, S&H purchases one drum of mineral spirits each month for dilution of the machine oil. Assuming a solvent density of 7 pounds per gallon, the emissions of this solvent amount to about 1,540 pounds or 0.77 tons per year.

There are 1,684 machine shops in L.A. County, 768 in Orange County and 66 in the cities of Riverside and San Bernardino. The total number of machine shops is 2,518. These are independent machine shops and the figures do not include captive machine shops that are part of a larger manufacturing facility. There are likely to be at least 500 additional captive machine shops. The total number of machine shops in the Basin is estimated at 3,018. Assuming that 40 percent of them use petroleum coolant and that their emissions are comparable to S&H Machine's emissions, the total VOC inventory in this category is 929 tons per year.

6.4. Vanishing Oil

Many oil supply companies offer products that are classified as vanishing oil. These oils are used in machines that are used for stamping and forming metal parts of various types. Vanishing oils do not have to be cleaned off after fabrication because they "vanish" or evaporate from the part. These vanishing oils have a large component of low vapor pressure solvent which is why they evaporate. The number of firms using vanishing oils in the Basin is not known but this category could be a large source of VOC emissions.

6.5. Minimum VOC Limits

More than 100 water-based cleaning formulations have been certified as Clean Air Solvents by the District. About two-thirds of these formulations have a VOC content of 25 grams per liter or less at the recommended dilution level. In the past, the District set the VOC content in certain categories, including general repair and maintenance cleaning, at 50 grams per liter because this was the detection limit for the analytic procedure. Improvements have been made in the analytic procedure so that a 25 gram per liter limit can be specified.

As an example of the reduction that can be achieved, consider the general repair and maintenance cleaning category. If the VOC limit were lowered from the current level of 50 grams per liter to 25 grams per liter, the District estimates that this change would achieve an emission reduction of 0.2 tons per day in 2004 and 0.23 tons per day in 2010. As another example, consider the Rule 1122 amendment for batch loaded cold cleaners and conveyorized cleaners that required the use of solvents with a VOC content of 50 grams per liter or less or the use of an airless/air-tight degreaser. Although this amendment resulted in a very large emissions reduction, of the order of 40 tons per day, the remaining inventory is significant, at 7.2 tons per day. By reducing the de minimus VOC content level from 50 to 25 grams per liter, an emissions reduction of 3.6 tons per day could be achieved.

In this section, the results of the project analysis are summarized and recommendations based on the project findings are made. During this project, IRTA worked with 10 facilities to test and analyze alternative low- or non-VOC cleaning methods. IRTA analyzed the results of the testing and compared the costs and feasibility of the high- and low- or non-VOC processes. IRTA also selected 11 additional facilities that had converted to low- or non-VOC alternatives in the past. In some of these cases, qualitative results were presented; in most cases, IRTA compared the costs and feasibility of the new low- or non-VOC process with the old high VOC process. In some of the case studies, companies had converted away from toxic solvents that are not VOCs.

Table 7-1 summarizes the categories of cleaning that were examined during this project. The current VOC content and vapor pressure limits are provided in the table. An estimate of the inventory for most of the categories is also given in the table. In some cases, the inventory is unknown.

In addition to the testing and/or analysis of alternatives performed for each of the categories in Table 7-1, IRTA also evaluated certain exemptions in the SCAQMD cleaning rules. Table 7-2 presents the exemptions that were analyzed.

In this section, the results and conclusions of the tests and analysis are presented for operations covered by Rule 1171, Rule 1122, Rule 1124, Rule 442 and other operations that are currently not covered by District rules. This section also summarizes IRTA's recommendations.

7.1. Results and Conclusions for Rule 1171 Cleaning Operations

7.1.1. Surface Preparation

IRTA worked with one company, Charles Caine, that uses handwipe solvents for surface preparation of parts prior to coating. Many companies have converted away from TCA and CFC-113 to VOC solvents for handwipe cleaning. With the exemption of acetone and the availability of water-based cleaners, most companies should be able to meet the 70 gram per liter VOC content cutoff level that is currently in the rule.

7.1.2. Product Manufacture--Medical Devices

IRTA worked with one medical device manufacturer on several applications that involved handwipe of components during manufacture. As Table 7-1 indicates, the VOC content level for this category is 900 grams per liter and 33 mm Hg. The reason the level is set with these values is that most companies use IPA for cleaning. In some cases, acetone proved to be an effective alternative and blends of IPA and D.I. water were suitable. In light of the complexity and diverse nature of the cleaning operations performed by medical device manufacturers, the low emissions inventory for this category and the FDA approval process required for changes, the VOC limits should not be reduced further.

7.1.3. Product Manufacture--Electronic Components

IRTA presented the results of an alternative used by Aerojet Electronic Systems in cleaning electronics, specifically printed circuit boards. Although Aerojet generally uses IPA for rework of their printed circuit boards, they do clean some of the boards that are reworked in their coveyorized water-based cleaning system. Printed circuit board rework is only one type of electronics operation that involves cleaning. Further work is needed in the electronics category to demonstrate cleaning alternatives before the VOC limits are changed.

7.1.4. Repair and Maintenance Cleaning--General

In the general repair and maintenance cleaning category, IRTA assisted two companies, Paul's Transmission and Newhall Carburetor, in converting to water-based cleaners to satisfy the January 1, 1999 deadline for 50 gram per liter cleaning materials. Paul's Transmission cleans valve bodies and Newhall Carburetor cleans and rebuilds carburetors

Table 7-2
Cleaning Rule Exemptions

of all kinds. These general repair and maintenance cleaning applications were thought to pose the greatest challenge for water-based cleaners. Both companies are satisfied with their water-based cleaners and both reduced their costs through the conversion. During the work with Paul's Transmission, IRTA investigated whether valve bodies with solenoids could be cleaned with water. The three that were tested all failed in the automobiles later. This confirms that a higher VOC limit for electrical assemblies is still needed for this industry. Another item that cannot be cleaned with water-based cleaners is the clutch plate friction materials that are bonded with an adhesive; the adhesive delaminates when it is placed in water. Aside from the two areas that require higher VOC limits or exemptions, the technology for reducing VOC emissions in this category is available today and cost effective.

IRTA worked with the Rocketdyne Division of Boeing on general field handwipe cleaning. Rocketdyne had already converted to a water-based cleaner for this purpose and the company agreed to test another aerosol water-based cleaner. The cleaner the company was using had a lower VOC content than the aerosol and it performed very well for cleaning all Rocketdyne's non-electrical field components. Many companies have already found acceptable low VOC alternatives. The exemption for field handwipe no longer appears to be necessary.

7.1.5. Repair and Maintenance Cleaning--Electrical Apparatus Components

IRTA also worked with Rocketdyne to test alternatives for cleaning electrical apparatus components which are energized. A few years ago, Rocketdyne stopped cleaning these components altogether because the company committed to stop using ozone depleting solvents. Solvents with flash points cannot be used to clean energized equipment. IRTA arranged for Rocketdyne to test four alternatives that did not have flash points. All four were found to be ineffective. At present, there appear to be no non-VOC alternatives and the only VOC alternatives that do not have flash points have toxicity problems.

7.1.6. Cleaning of Coating or Adhesive Application Equipment

IRTA tested alternatives in cleaning of coating application equipment with three companies, Steelcase, Charles Caine and Hydro-Aire. The current VOC content limit in the category is 950 grams per liter and there is also a vapor pressure limit of 35 mm Hg. For waterborne coatings, the Charles Caine case study indicated that water is a suitable cleaning agent. IRTA analyzed Steelcase's conversion to waterborne adhesives and water for cleaning the adhesive application equipment. The water works effectively for that purpose. For solventborne coatings, in some cases, a water-based cleaner cleaned effectively but it foamed. Acetone cleaned most of the coatings well and blends of acetone with the solvent currently used for cleaning also worked well. The results are promising and indicate that, with some further testing, it is likely that low- or non-VOC solvents can be identified for cleaning solventborne coatings in application equipment.

7.1.7. Cleaning of Ink Application Equipment

IRTA worked with and analyzed the operations of three companies, R R Donnelley & Sons, San Bernardino Sun and Nelson Name Plate, that use cleanup solvents for lithographic printing presses. The majority of the solvent at R R Donnelley is used for cleaning parts that are removed from the press. Some of the cleaning at the San Bernardino Sun is for off-press parts. All of these parts can be cleaned effectively with water-based cleaners. The San Bernardino Sun tested a water-based cleaner for cleaning rollers on the press and it performed effectively but was expensive. Nelson Name Plate has been using an acetone-based formulation with about 10 percent VOC as a blanket and roller wash for at least two years. Although water-based cleaners proved successful for some of the operations in this study, the facilities that participated in the testing do not span the range of printing operations in the Basin. Additional tests and demonstrations are required to set a lower VOC limit.

IRTA worked with Nelson Name Plate on a screen printing operation. The company converted to a water-based process for recycling the screens several years ago. IRTA and Nelson tested alternatives for cleaning the screens in between runs. A 50 percent acetone blend with the VOC solvent the company uses currently was found to be very effective. The results of the work with Nelson are promising but more testing of different alternatives in additional screen printing operations is required.

7.2. Results and Conclusions for Rule 1122 and NESHAP Cleaning Operations

7.2.1. Batch Loaded Cold Cleaners (BLCCs)

This section presents the results of three BLCC conversions. In one case, S&H Machine, Inc., IRTA assisted the company in converting to water-based cleaners from mineral spirits. Rule 1122 specifies that solvents used in BLCCs must have a VOC content of 50 grams per liter or less by January 1, 1999. Another option is for firms to convert to an airless/air-tight degreaser. S&H Machine converted to water-based cleaners and has found them to be effective. The company reduced the cleaning cost slightly through the conversion.

In another case, Hydro-Aire converted four BLCCs that used mineral spirits to water-based cleaners. This company increased their costs substantially through the conversion but found the water cleaners very effective. The company's costs increased because the cost for purchasing the cleaner was higher and the company decided to have the cleaning operations serviced.

IRTA also worked with Litton Guidance & Control Systems, a company that manufactures laser optics. The company converted from a variety of VOC solvents used in BLCCs to water-based cleaners. Litton reduced their costs significantly through the conversion. The company is covered by an exemption in Rule 1122 which allows the continued use of high VOC solvents until January 1, 2003. Instead of relying on this exemption, Litton decided to convert now.

The results of the conversions at S&H Machine and Hydro-Aire suggest that using water-based cleaners and in place of solvents in BLCCs is well demonstrated in machine shops and further problems are not expected. The Litton conversion demonstrates that the exemption for specific applications in Rule 1122 may no longer be necessary.

7.2.2. Electronics Cleaning

Results for two companies, Hydro-Aire and Aerojet Electronic Systems, are presented as conversions in printed circuit board defluxing. In the past both companies used ozone depleting solvents for defluxing printed circuit boards in an assembly process. Aerojet made the conversion several years ago and IRTA assisted Hydro-Aire in the conversion during this project. The costs are not available for Aerojet but Hydro-Aire reduced their costs in the conversion. Water-Based cleaners have been used for the last several years by many firms in the Basin for printed circuit board defluxing and that process and the use of low solids flux is likely to be suitable for virtually all operations. The low solids fluxes do not require cleaning because they leave only small residues on the boards.

7.2.3. General Metal Vapor Degreasing

IRTA worked with four companies, California Electroplating Inc., Hydro-Aire, Nelson Name Plate and Barranca to assist them in converting from NESHAP solvents in vapor degreasing to water-based cleaners for general metal cleaning applications. All of the companies found the water-based cleaners effective for cleaning and all reduced their costs through the conversion.

7.2.4. Precision Cleaning

During this project, IRTA worked with two companies to assist them in converting to water-based cleaning in precision cleaning operations. One of the companies, Astro Pak, used a VOC solvent, n-propyl bromide, and converted to a water-based cleaner. The water-based cleaner was more effective than the solvent in cleaning the company's parts. Another company, Kaiser Electroprecision, did limited testing of an alternative but has to do additional work to verify and perhaps improve the purity of their D.I. water.

7.2.5. Conclusions

The information presented here indicates that water-based cleaners are viable substitutes for vapor degreasing in general metal cleaning, electronics cleaning and precision cleaning. Virtually all vapor degreasing operations could be converted to water-based cleaners.

7.3. Results and Conclusions for Rule 1124 Cleaning Operations

IRTA worked with one company, Hydro-Aire, during the project on handwipe operations that are covered by Rule 1124. In these cleaning applications, acetone and other non-VOC materials proved effective. IRTA also reported on a conversion that was made several years ago by Rocketdyne. That company converted to water-based cleaners, an exempt HFC and cyclohexane which is a VOC for their cleanliness verification handwipe operations.

7.4. Results and Conclusions for Rule 442 Operations

During this project, IRTA worked with one company that has an operation that falls under District Rule 442. IRTA assisted Nelson Name Plate in adopting a low VOC stripping agent containing acetone. IRTA also identified two large Rule 442 emitters. These include MCP Foods and Stewart Film Screen.

7.5. Results and Conclusions for Other Cleaning Operations

In the course of this project, IRTA identified several other promising areas where the District could achieve VOC reductions. In the first area, cleanroom maintenance, IRTA has extensively tested alternatives. Emissions from janitorial/institutional cleaning, the second area, are likely very high. A third area IRTA identified is solvent use for dilution of machine coolants; emissions from these sources are also very high. The fourth area is the use of vanishing oil which contains solvents that are emitted. The inventory for this use is unknown. Finally, the fifth area is a reduced VOC limit for applications where water-based cleaners are used. The District has determined that sufficient accuracy can be achieved with the limit at 25 rather than 50 grams per liter.

7.6. Recommendations

IRTA tested alternatives in several cleaning application areas where the VOC emissions limits are still very high during this project. IRTA also identified several other areas with high VOC emissions that are not regulated today. IRTA has defined four categories for classifying the additional VOC reductions that could be required in the future. The categories are:
o No Further Action
o Short-Term Reductions
o Medium-Term Reductions
o Long-Term Reductions

The category "no further action" includes areas where there are so many technical or institutional difficulties that further VOC reductions would not be justified. Institutional difficulties could be encountered when an industry is regulated by other entities that may have higher priority than an air regulatory agency. These would include the FDA regulation of the medical industry or the military specifications that affect some aerospace operations. The emission inventory for these areas is also small so that further regulation is simply not worthwhile.

The category "short-term reductions" includes areas where there are not technical or institutional barriers to requiring a VOC reduction over the next few years.

The category "medium-term reductions" includes areas that show great promise for large emission reductions but which raise issues that need to be resolved. In some cases, more technical work is required to determine what VOC content levels are appropriate. In other cases, certain institutional constraints must be overcome before the area can be further regulated. In still other cases, both technical and institutional work is still required.

The category "long-term reductions" includes areas that are very promising for future regulation but the technologies for achieving the emission reduction are currently not known. In these cases, substantial additional technical work would be required to determine appropriate VOC limits.

The areas that fall into each of the categories are listed below and an explanation for the categorization is provided.

No Further Action

o Cleaning During Product Manufacture--Medical Devices. The VOC emissions inventory for this category is very small and all changes require FDA approval. The FDA approval process is onerous and very time consuming.

o Repair and Maintenance Cleaning--Medical Devices. The same reasoning as for product manufacture cleaning applies here.

o Repair and Maintenance Cleaning--Energized Electrical Apparatus Components. Most companies are currently using HCFC-141b which will be banned in 2003. VOC solvents generally have flash points and solvents with flash points cannot be used in this application. The VOC emissions inventory is likely small.

Short-Term Reductions

The VOC reductions discussed in this category can be taken immediately. In some cases, IRTA recommends that exemptions be kept until they expire. Additional work is not required to further regulate.

o Rule 1122 Vapor Degreasers and NESHAP Regulation.

The District could regulate vapor degreasers the same way BLCCs and conveyorized cold cleaners are currently regulated. The options are to use a solvent with a VOC content of 50 grams per liter or less or to use an airless/air-tight degreaser. The information presented here on the conversions firms have made in general metal cleaning, electronics and precision cleaning indicate that water-based cleaners can be used in virtually all vapor degreaser applications to replace VOC solvents. For users that wish to continue to use VOCs, the airless/air-tight degreaser option is available.

NESHAP solvent operations could be handled in the same way. Companies using these toxic solvents could be required to convert to solvents with a VOC content of 50 grams per liter or less. Alternatively, if they wished to continue using the NESHAP solvents, they could use them in an airless/air-tight degreaser.

Rule 1122 contains an exemption for vapor degreasers and BLCCs that are 1 square foot in surface area that are used to clean various types of components. The Litton Guidance & Control Systems case study indicates that users of BLCCs for cleaning optics and other precision equipment can convert to non-VOC alternatives. This exemption is no longer necessary. The District could modify the rule to require all cleaning equipment (including vapor degreasers) to use solvents with a VOC content of 50 grams per liter or to use an airless/air-tight degreaser when the exemption expires. At this stage, the 50 gram per liter limit or the airless/airtight degreaser option will apply to all equipment regulated in Rule 1122. If there are still a few areas which require VOC solvents, narrow specific exemptions could be allowed.

o Cleanroom Maintenance--non-biocidal

The study IRTA performed several years ago and the information presented here lead to the conclusion that plain D.I. water can be used for all cleanroom maintenance where biocidal control is not required. The District could immediately regulate this category.

o Field Handwipe Repair and Maintenance Cleaning

Rule 1171 includes an exemption for field handwipe repair and maintenance cleaning. The work with Rocketdyne and information from other users indicates that this exemption is no longer needed. It should be allowed to expire on January 1, 2001.

o Coating and Adhesive Application Equipment--Cleanup of Waterborne Coatings and Adhesives

The rule limit for this category should be set at 50 grams per liter. Plain water is generally suitable for cleanup of waterborne coatings and adhesives.

o 50 Grams per Liter VOC Level

The sensitivity of the District laboratory analysis has improved. The de minimus level of 50 grams per liter specified in many categories could be changed to 25 grams per liter.

Medium-Term Reductions

The reductions discussed in this category require further work. In some cases, more technical work is required. In other cases, institutional constraints need to be dealt with. In a few instances, both technical work and resolution of institutional constraints are necessary.

o Cleaning of Ink Application Equipment

The results of the conversions and testing presented here for lithographic and screen printing are extremely promising. They indicate that components that can be removed from the press can be cleaned with water-based cleaners. They also indicate that some presses could be cleaned with water-based cleaners. Technical issues remain to be resolved however. Additional research and testing is needed in all areas of printing to determine appropriate future VOC levels.

o Coating and Adhesives Application Equipment--Cleanup of Solventborne Coatings and Adhesives

The results of the testing during this project are very promising. Many coatings can be cleaned with water-based cleaners. A technical drawback was that the water-based cleaner selected for testing foamed. Acetone may be suitable for cleanup of a significant portion of the coatings used today. Technical issues remain to be resolved, however. Additional research and testing is needed on all kinds of coatings to determine what cleaning agents would be suitable. This would allow the District to determine the appropriate VOC levels for cleaning all types of coating and adhesive application equipment.

o Janitorial/Institutional Cleaning--Non-Biocidal, Non-Sterile

This category is not currently regulated. It is not clear whether CARB has jurisdiction over this type of cleaning. If CARB does have the authority, the District could work with CARB to regulate this category. Institutional issues require resolution. The VOC emission inventory in this category is likely to be very high.

o Janitorial/Institutional Cleaning--Biocidal, Sterile

This category has the same institutional issues as the category above. For hospitals and other medical institutions, this type of cleaning may be regulated by the FDA which would pose additional institutional constraints. There may also be technical issues if VOC solvents are used to achieve the biocidal or sterile conditions. Technical work would be needed to identify non-VOC biocidal or sterile agents that would be effective.

o Cleanroom Maintenance--Biocidal Control

Additional technical work is required in this category. Non-VOC biocidal agents need to be identified and tested to ensure they are effective. The FDA approval process is likely to be long and difficult and this is an institutional constraint that requires further work.

o Electronic Apparatus & Components--Product Cleaning and Surface Preparation and Repair and Maintenance Cleaning--Non-Energized

Additional technical work is required in this area. Water-based cleaners and exempt solvents are likely to prove feasible for cleaning electronic components of all kinds.

o Machine Oil Dilution

More technical work is required in this category. Additional research on the types of machines available to machine shops and the oil they use is required. Testing may be required to determine if synthetic oils can be used in all types of machines. The VOC emissions inventory in this category is very large.

Long-Term Reductions

The cleaning applications in this category require substantial further work before further VOC reductions could be required. Generally, the technologies for reducing the VOC emissions are unknown presently.

o Rule 1124 Handwipe Cleaning

Some of the handwipe operations covered by Rule 1124 could be fairly easily converted to water-based or acetone alternatives. These would have to be identified and distinguished from the more difficult applications. Alternatives in aerospace handwipe operations must generally be pursued on a case-by-case basis. Additional testing and research would be required before this category could be further regulated. It is worth investigating, however, because of the very large VOC emission inventory.

o Rule 442

Emissions from the operations that fall under this rule are likely to be very high. This category is worth pursuing for this reason. The operations covered by Rule 442 would have to be identified and classified. Additional research and testing might be required to identify suitable low- and non-VOC alternatives.

Table 7-3 summarizes the recommendations discussed in this section.

VIII. BIBLIOGRAPHY

K. Verheyen, B. Ekstrand, K. Wolf, M. Morris, R. Pease, J. Quitmeyer and J. Fuchs, "Precision Contract Cleaning Company Converts to Aqueous Process," to be published in Precision Cleaning Magazine, October 1999.

K. Wolf and M. Morris, "Alternatives to Solvents in Cleaning Applications: Case Studies of Small Business Conversions to Water-Based Cleaners, prepared by the Institute for Research and Technical Assistance for EPA, June, 1999.

"The Alternative," published by the Institute for Research and Technical Assistance, Spring 1999.

M. Morris and K. Wolf, "Water-Based Repair and Maintenance Cleaning: Case Study Conversions," prepared by the Institute for Research and Technical Assistance for Southern California Edison, December 18, 1998.

M. Morris and K. Wolf, "Water-Based Parts Washer Systems: Case Study Conversions, prepared by the Institute for Research and Technical Assistance for EPA and Santa Barbara County APCD. December 11, 1998.

S. Greenberg, B. Lamb, R. Johnson, G. Robertson and J. Topalian, "New Paradigms in Environmental Compliance: A Case Study of Progress by an Aerospace Contractor," Federal Facilities Environmental Journal, Winter 1998, p. 25.

"The Alternative," published by the Institute for Research and Technical Assistance, Spring 1998.

"The Alternative," published by the Institute for Research and Technical Assistance, Spring 1996.

M. Price, "Solvent Handwipe in Assembly Areas," RSS-8944, Rocketdyne Division of Boeing, Section 6, p. 6-1.

V. Douglas and M. Fritzemeier, "Alternate Cleanliness Verification Techniques," RSS-8944, Rocketdyne Division of Boeing, Section 12, p. 12-1.

APPENDIX B

Stand-Alone Case Studies for Selected Companies

TRANSMISSION SHOP CONVERTS TO WATER-BASED CLEANING PROCESS
Avoids Rusting Problems

Paul's Transmission is a family business with two locations, one in Santa Monica and one in Culver City. The Santa Monica shop has five employees and the Culver City shop has six. The company repairs and rebuilds transmissions of all kinds. In some cases, the transmissions are stored for a year or more before they are installed.

For many years, Paul's Transmission used two mineral spirits sink-on-a-drum parts cleaners in each shop to satisfy their cleaning needs. The parts cleaners were supplied by a service which also arranged for changeout of the baths and disposal. Like most transmission shops, the company also relied on a spray cabinet that used a water-based cleaner for cleaning most transmission parts other than valve bodies.

In 1996, the South Coast Air Quality Management District (SCAQMD) modified their cleaning rule, Rule 1171, which applies to repair and maintenance cleaning. It required a conversion to water-based cleaners by January 1, 1999. In 1998, as part of an EPA project, IRTA began work with Paul's Transmission, which would be affected by the rule requirements.

IRTA arranged for Paul's Transmission to test a water-based sink-on-a-drum and an ultrasonic cleaning system. The ultrasonic cleaning system proved to be very effective for cleaning the valve bodies and other parts which have many passages that are difficult to clean with traditional methods.

After a few months of testing, Paul's Transmission purchased a Gray Mills sink-on-a-drum and a filtration unit supplied by Applied Cleaning Technologies (ACT) and an 18-inch cube ultrasonic cleaning system made by Alpha Cleaning Systems. "The ultrasonic system is the best system I've found for cleaning the valve bodies," says Vance Griffitts, owner of Paul's Transmission. "We were concerned about rusting especially for the valve bodies we store for a long time. We figured out we could soak them in transmission fluid after cleaning to prevent rusting," he says.

The water cleaning baths are changed out about every two months. "We put the spent cleaners in the spray cabinet," says Mr. Griffitts. "It gives it some zip." The shop has always used a clarifier to dispose of the spent water-based cleaner from the spray cabinet so there are no additional disposal costs from use of the new water cleaning systems.

The shop cleans some valve bodies with solenoids and electronic switches which are classified as electrical components. SCAQMD Rule 1171 includes an exemption for electronics cleaning. Paul's Transmission kept one of the mineral spirits parts cleaners to clean these electronic parts.

"I didn't believe the water cleaners would work at first," says Mr. Griffitts. "The workers like the new parts cleaner because it doesn't damage their hands. The new cleaning system costs are lower than the cost for solvent cleaning. "We like it when we can help the environment and save money at the same time," he says.

CARBURETOR REBUILDER MOVES TO WATER-BASED CLEANING
Water-Based Cleaners Effective for Internal Passages

Newhall Carburetor & Auto Repair is located outside Los Angeles in Newhall. Jay Tallent, the owner of the shop, does all the rebuilding and repair work himself. He has one employee who handles the bookkeeping. Mr. Tallent started working in the shop in 1991 and purchased it in 1993. He has repaired and rebuilt thousands of carburetors over the period.

Like other auto repair shops, Newhall Carburetor relied on a mineral spirits parts cleaner for quickly cleaning parts of all kinds. About three years ago, Mr. Tallent decided to investigate water-based parts cleaners. After evaluating the systems available at the time, he decided to purchase a Zymo enzyme system. The enzyme system cleans the parts well and minimizes disposal costs. Says Mr. Tallent, "the enzyme unit performs about as well as the mineral spirits unit and it is much cheaper. It's better for your hands and the environment."

The shop also had a carburetor cleaning unit that used a mineral spirit combined with a chlorinated solvent for cleaning the carburetors. Although the carburetor cleaning unit used agitation to enhance the cleaning, an hour of cleaning time was required for each carburetor. An additional 20 minutes of hand cleaning in the enzyme system was also necessary to clean the carburetors adequately.

In a project sponsored by Southern California Edison, EPA and the Pollution Prevention Center, IRTA worked with Newhall Carburetor in 1997 to test alternative methods of cleaning carburetors with water-based cleaners. At first, a heated carburetor cleaning unit was built specifically for this application and tested at the shop. It did not perform well even though the water cleaner was heated to a fairly high temperature. A second unit, an ultrasonic system from Alpha Cleaning Systems, was then tested and it performed very well. After a period of testing, Mr. Tallent decided to purchase an 18-inch cubed ultrasonic system. He is using it very successfully today for cleaning all the carburetors the shop repairs.

Mr. Tallent is extremely pleased with the aggressive cleaning action of the ultrasonic system and its ability to penetrate the blind holes and crevices in carburetors. "I had a $900 boat carburetor that was caked with varnish that I wouldn't have been able to remove with traditional cleaning methods. In the past, I would have had to discard the carburetor," he says. "I was able to clean it up really well with the ultrasonic cleaner in about 30 minutes. It saved me a lot of time and the customers didn't have to purchase a replacement carburetor."

The ultrasonic unit reduces Mr. Tallent's labor significantly. "I save a lot of time with this system. All I do is turn the unit on and I can walk away and do other things," he states. Because of the labor savings, the total cost of using the new ultrasonic system is only about one-third the cost of using the old carburetor cleaning unit.

S&H Machine, Inc. is a small machine shop with 13 employees located in Burbank. The company is a job shop that machines parts made of aluminum, stainless steel and carbon steel. Many of the parts are complex with blind holes and threads. S&H Machine's primary customers are aerospace subcontractors.

For many years, the company used mineral spirits for cleaning the parts. In one side of the shop, there are 21 stations where operators machine parts. After the parts are machined, the workers clean them in coffee cans that contain solvent. The parts are allowed to soak for a time, the worker removes them from the coffee can and blows off the excess solvent with compressed air back into the machines.

In another side of the shop, two 5-gallon batch loaded cold cleaners were used to final clean many of the parts. The parts are dipped in the tank and allowed to soak for a time. The worker removes them from the tank and blows them off individually to remove excess solvent.

In 1998, IRTA began work with S&H Machine to assist them in converting from mineral spirits to water-based cleaning processes. The South Coast Air Quality Management District (SCAQMD) had modified their cleaning rule, Rule 1122. This rule, which covers solvents used in batch loaded cold cleaners, required companies to use solvents in these types of cleaning units with a VOC content of 50 grams per liter or less or to convert to an airless or airtight degreaser by January 1, 1999.

S&H Machine was affected by the rule because their coffee cans and 5-gallon containers are classified as batch loaded cold cleaners. Because airless/airtight degreasers are very expensive, S&H Machine decided they would work to implement water cleaning processes. "We were very skeptical of water-based cleaning at first," says David Fisher, manager and Vice President of S&H Machine. "We thought water cleaners would rust the parts and we didn't think they would clean as well."

IRTA arranged for S&H Machine to test two water-based cleaning units. One was a 30-gallon Gray Mills sink-on-a-drum supplied by Applied Cleaning Technologies (ACT) which was intended to substitute for the coffee cans. The other was a small ultrasonic cleaning unit made by Western Sonics which was used to test the parts cleaned in the 5-gallon containers for final clean.

After a few months of successful testing, S&H Machine decided to purchase the Gray Mills unit. They also purchased seven additional 15-gallon Gray Mills sink-on-a-drum units for use at the 21 work stations. For the final clean area, the company purchased a two-bath ultrasonic unit. The wash and rinse baths each hold 30 gallons.

IRTA and S&H Machine tested a few different water-based cleaning formulations to identify one that was suitable for removing all of the oils used by the company and encountered in the cleaning process. IRTA encouraged the company to use one cleaning formulation throughout for the parts cleaners and the ultrasonic cleaning unit. That way, when the ultrasonic wash bath required changeout, it could be used in the parts cleaners which could tolerate dirtier cleaning agent. The company settled on a neutral cleaner called Power Clean Scrub Tub made by ACT. "We needed a neutral cleaner for both applications so the workers' hands would not be damaged in the parts cleaners," says David Fisher.

The Gray Mills unit has been operating at S&H Machine now for six months and it has not yet required changeout. The ultrasonic unit which is used for final clean will probably require changeout more often, perhaps every two months.

Mr. Fisher has been persuaded that water-based cleaning is effective for S&H Machine's parts. "We like the new water cleaning systems. They are better for the workers than the solvent and better for the environment," he says. "The surprise is that the water-based cleaners are slightly less costly than the solvents."

Litton Guidance & Control Systems is located in Woodland Hills, California. The company makes laser-based guidance systems for space applications. The optical components must meet stringent performance specifications and cleaning is a major part of the operation.

The company historically used ozone depleting solvents, CFC-113 and 1,1,1-trichloroethane (TCA), for their cleaning. Litton began work several years ago on alternatives when the production bans were announced. All of their operations were converted away from CFC-113 and TCA, primarily to VOC solvents and water-based cleaners with high concentrations of VOC solvents.

The South Coast Air Quality Management District (SCAQMD) amended Rule 1122 "Solvent Degreasers" in July of 1997. The amendments affected VOC solvents that are used in batch loaded cold cleaning operations. The rule requires companies to use solvents with a VOC content of 50 grams per liter or less or to use the higher VOC solvents in an airless or airtight degreaser beginning in January 1999. Since Litton had many operations using VOC solvents, they were strongly affected by the rule.

IRTA began work with Litton in 1998 to assist the company in evaluating their processes and in adopting low- and non-VOC solvents so they could comply with the January, 1999 deadline. Says Gary Augeri, Member of the Technical Staff at Litton, "our operations might have been covered by one of the exemptions in Rule 1122 so we could have continued to use the VOC solvents. Litton Manufacturing Management wanted to set an example and we decided to make a commitment to switch away from these solvents."

At this stage, Litton Optics Manufacturing has converted virtually all of their cleaning processes away from VOC cleaners in the frame, substrate and prism operations in the optics shop. For frame manufacture, wax was used to plug the frame bores to prevent lapping compound from entering the internal bores. Litton eliminated a cleaning step that employed n-methyl pyrrolidone (NMP) by using plugs with O-rings to block the frame bores as a physical barrier to the lapping compound. The lips of the plugs are now sealed with adhesives which are removed with a Liquinox detergent. Epoxy is used to bond the frames to holding fixtures during lapping and polishing. In the past, NMP was used to remove this epoxy. Very hot detergent is now used to separate the frame from the fixture. The thermal expansion difference between the glass part and the metal fixture causes the debonding.

In the substrate operation, pitch was used to hold the mirror substrates to mounting blocks during lapping operations. NMP, methanol and methylene chloride were used in the past for cleaning. Litton now uses thermoplastic instead of pitch for this bonding. Acetone is currently used to dissolve most of the thermoplastic; this is followed by a soak in an Armakleen detergent made by Church & Dwight.

In the prism operation, wax is used to bond the prisms to mounting blocks for lapping and polishing. A terpene-based cleaning process was used to dissolve the wax and clean the parts. Litton has converted to Daraclean 121, a water-based cleaner made by W.R. Grace for this cleaning process.

All of the parts are put through a final clean either with hot water alone or with hot water and detergent. In some cases, ultrasonics are necessary to achieve the required cleanliness.

"The new processes work very well," says Mr. Augeri. "In some cases, we were able to use different materials in our processes and could avoid cleaning altogether. In other cases, we could substitute water-based cleaners. We found we don't have to rely on solvents for getting the cleanliness we need. The new water-based cleaners are better for the environment and for our workers."

California Electroplating Inc., a small company with 19 employees located in Los Angeles, began operation in 1915. The job shop offers chromium, copper and nickel plating finishes for zinc die cast, steel and brass stock.

The company has two plating lines. The first line is a computer-controlled automatic rack plating line. The parts that are plated on this line contain oil and various other types of contaminants but do not generally contain polishing or buffing compound. The second line, the hand line, is used for plating smaller volume custom jobs. The parts on this line contain buffing compound.

IRTA began work with California Electroplating in late 1997. The company wanted to find a suitable alternative to the solvents they used to clean the parts prior to plating. For many years, the company used 1,1,1-trichloroethane (TCA) and perchloroethylene (PERC) in a vapor degreaser for this purpose. Production of TCA has been banned because the chemical causes ozone depletion and PERC is classified as a toxic.

IRTA performed field testing in an ultrasonic cleaning system with a water-based cleaner that is especially effective for removing buffing compound. The cleaner, Daraclean 236, is supplied to California Electroplating by Applied Cleaning Technologies (ACT). "I was skeptical at first that a water-based cleaner could remove our polishing compound," says Frank Grana, one of the owners of California Electroplating. "The cleaner did a good job on all our parts and we decided to purchase an ultrasonic system."

At this stage, the company is using the new ultrasonic cleaning system which was built by Sonicore to clean all the parts from the hand line. Mr. Grana plans to install a hoist over the next few months so the cleaning system can also be used for the parts on the automated line which must be cleaned in baskets. "We're saving a lot in labor cost because the ultrasonic system is so effective on the polishing compound. We want to shut down the vapor degreaser entirely."

California Electroplating is able to use the cleaning bath for two months before it requires changeout. The company has experimented with the proper concentration and found that eight percent is optimal. "The cleaner is expensive but we use a low concentration and the bath life is much longer than I expected," says Mr. Grana.

Mr. Grana cautions that equipment manufacturers that make ultrasonic systems for platers should design them to be more hardy. Moisture ended up penetrating the generator, which is part of the cleaning system. Mr. Grana solved the problem by placing a light, which dries the sensitive components, inside the generator casing. "We've gone through four printed circuit boards since we started operating the ultrasonic cleaner," says Mr. Grana. "They fail quickly because of the corrosive environment in the plating shop," Even so, he is very satisfied with the new system. "We would not be able to clean as quickly or as well without the unit," he says.

Annual Cost Comparison for California Electroplating

Hydro-Aire is a division of Crane located in Burbank, California. The company has 572 employees. Hydro-Aire manufactures braking systems, pumps and air locking devices and is a Boeing subcontractor. Hydro-Aire also repairs the pumps used in military and commercial aircraft like the C-17 and the C-130 transport.

IRTA began work with Hydro-Aire in 1997 to assist the company in converting away from solvents in a variety of applications. An earlier article in the Spring 1998 Alternative described the company's conversion away from 1,1,1-trichloroethane (TCA) in a printed circuit board assembly process. This article focuses on Hydro-Aire's conversion away from TCA in vapor degreasing and mineral spirits in batch loaded cold cleaning.

For many years, Hydro-Aire used TCA in four vapor degreasers for cleaning parts in three areas of the plant including the deburring area, the hone, lap and grind area and the non-destructive testing area. Production of TCA was banned worldwide in January, 1996 because the chemical contributes to stratospheric ozone depletion. Although inventories are still available, the chemical, at this stage, is very expensive. IRTA began working with Hydro-Aire to identify an alternative water-based cleaning process that would be suitable for cleaning the company's parts. "We needed a cost effective alternative and one that would be better for the workers and the environment," says Tommy Jennings, Environmental Manager at Hydro-Aire.

Hydro-Aire's parts are made of many materials including stainless steel, aluminum, steel and titanium. Laboratory testing identified water-based cleaners that would be suitable for cleaning the range of contaminants that must be removed from the parts. Extensive field testing revealed that an ultrasonic system would be required for the cleaning because Hydro-Aire's parts had many passages, blind holes and crevices. After considerable evaluation of various types of systems, the company decided to purchase three ultrasonic cleaning systems from Western Sonics. These systems each have a wash bath, two rinse baths and a hot forced air dryer. The wash and two rinse baths contain ultrasonics. All three systems are automated with a robot that transfers the baskets from station to station.

Two of the ultrasonic systems have been installed and are up and operating in the deburring area and the hone, lap and grind area. The third system is scheduled for installation shortly. In all three systems, a rust inhibitor is used to supply a protective coating for the ferrous metal parts in the first rinse. Deionized water is used in the second rinse and these rinses are routed to a closed loop water recycling system which removes the soils and returns clean water to the system. "We programmed the systems to go through a particular sequence, depending on the type of part being cleaned and the cycle that's required," says Jon Zavadil from Western Sonics.

The water-based cleaners that are being used in the ultrasonic systems are made by W.R. Grace and are supplied to Hydro-Aire by Applied Cleaning Technologies (ACT). "The new cleaners are environmentally friendly and they do a very good job on the parts," says Mike Halbert of ACT.

IRTA also worked with Hydro-Aire to find an alternative to three batch loaded cold cleaning units that used mineral spirits. South Coast Air Quality Management District (SCAQMD) Rule 1122 "Solvent Degreasers" requires batch loaded cold cleaners to use solvents with a VOC content of 50 grams per liter or less or to be replaced with an airless/airtight degreaser by January 1, 1999. Airless/airtight degreasers are very expensive and it would not have been practical to use this option in Hydro-Aire's machine shop.

IRTA arranged for Hydro-Aire to test water-based sink-on-a-drum units as replacements for the batch loaded cold cleaners. The company purchased three of these systems from ACT. The cleaner used in the systems, also supplied by ACT, is a neutral chemistry which does not damage the workers' hands.

"We're very happy with the new water-based cleaning systems," says Mr. Jennings who engineered and worked very hard on the conversions. "When we used the vapor degreasers, our annual costs were close to $200,000. Even though Hydro-Aire had to purchase the ultrasonic systems, our costs have been reduced to about $70,000 per year." We improved the cleaning operations for the workers and the environment and we also saved a lot of money."

Nelson Name Plate, a small firm in Los Angeles with 200 employees, was founded in 1946. The company manufactures metal and plastic name plates and makes membrane switches for many industries, including the medical industry. The firm fills about 10,000 orders for customized name plates each year.

Until recently, Nelson used about 4,000 gallons of 1,1,1-trichloroethane (TCA) annually for cleaning the stock used to make the name plates. Metal sheets arrive at the facility with a light coating of oil. Most of the sheets are 18 inches by 24 inches but sheets that are 12 inches by 40 inches are also processed. The average thickness of the sheets is 0.02 inches. Ninety percent of the stock is aluminum, 7 percent is stainless steel and 3 percent is brass. About 1,000 metal sheets were cleaned in the 50 gallon vapor degreaser each day.

Under one of IRTA's Pollution Prevention Center projects, IRTA staff began working with Nelson to identify, test and implement an alternative to TCA. Production of the solvent was halted worldwide on January 1, 1996. After extensive testing in the laboratory and at vendor facilities, a suitable process was identified. Cleaning the brass stock proved especially challenging.

Nelson purchased a conveyorized cleaning unit from Westek, an equipment manufacturer in Arcadia, California. The machine was custom designed and includes a wash section, an isolation chamber, two rinse sections and a dryer. An oil skimmer is used to remove the excess oil from the wash bath.

The water in the isolation chamber is plumbed to the sewer. Because of other plant operations, Nelson already had a sewer discharge permit. The Bureau of Sanitation indicated that no modification to the permit was necessary. The rinses use deionized water which is produced through a closed loop system. Says Jim Banis of Westek, "It is virtually always cost effective to close loop the process when deionized water is used for the rinse. The closed loop system eliminates the need for a constant hot deionized water flow to the final rinse. This type of system provides a payback in a short period of time."

Nelson is using a cleaning agent supplied by W.R. Grace. The cleaner has been formulated to reject oil which is removed by the skimmer. The cleaning agent contains less than two percent VOC so an air district permit is not required.

The water cleaning operation eliminated the use of 4,000 gallons of TCA annually. IRTA is assisting Nelson in converting away from two other operations that currently use TCA. Once the conversions are complete, Nelson's emissions will be under the threshold for obtaining a Title V Operating Permit.

Tom Cassutt, President of Nelson, is very pleased with the new water cleaning system. "The results of our process change have been all positive. The savings we realize from reduced solvent usage will pay for our initial capital outlay in less than two years. The new water cleaning system is more environmentally friendly, and our parts are cleaned more thoroughly today than with our old vapor degreaser."

Dean Delahaut purchased Barranca Diamond Products in October, 1998. The company, which has seven employees, manufactures diamond saw blades which are used for cutting in a variety of industries.

The previous company owner used a perchloroethylene (PERC) vapor degreaser in a cleaning operation. At the time Mr. Delahaut purchased the company, the South Coast Air Quality Management District (SCAQMD) permit on the vapor degreaser had lapsed. On July 10, SCAQMD Rule 1401 was amended to include PERC on the list of toxics. This rule regulates new and modified sources of toxics. Since there was no permit on the PERC vapor degreaser, it would be classified as a new source of toxics. When Mr. Delahaut inquired about obtaining a permit, he was informed that, because of the risk posed by the PERC emissions to the surrounding community, it was not likely that he could use the solvent.

At that stage, IRTA began working with Barranca to identify, test and implement a suitable water-based cleaning alternative. Barranca produces between 50 and 100 saw blades per day. The company receives the cold rolled steel in the form of flat stock. The parts are first notched and a metallurgical powder is added. They are then sintered in an electric furnace at 1700 degrees F. The parts are cooled overnight and are put through a grinding operation the following day. During the process, the blades are cleaned twice. After the final cleaning, the last step in the process at Barranca, they are sent outside to be painted.

IRTA recommended that Barranca send a set of parts to a cleaning laboratory. Laboratory testing identified a few cleaners that might be effective. IRTA arranged for Barranca to use a Hydro-Blast spray cabinet supplied by Applied Cleaning Technologies (ACT) for testing the best cleaning agent. The cleaner that was selected for field testing is Daraclean 200 which is made by W.R. Grace and is supplied by ACT. It is an alkaline cleaner that contains a corrosion inhibitor to protect the steel from rusting. After a few weeks of field testing, it was determined that the blades did not require a rinse. Says Mr. Delahaut, "I was surprised at how well the water-based cleaning process worked. There was less rusting on the parts than when they were cleaned with PERC."

Barranca decided to purchase the spray cabinet that was used at their site for the field testing. The price of the equipment was only $9,000. Says Mike Halbert of ACT, "I had the spray cabinet Barranca needed on the floor of my test center. Barranca is a small company and I could offer it at a discount." If Mr. Delahaut had decided to continue using PERC, the company would have needed a new airless/airtight degreaser that carries a cost of about $200,000.

The new system has been operating efficiently since November, 1998. In January, after two months of operation, the bath was spent and required changeout. IRTA assisted Barranca in getting permission to discharge the spent cleaner as long as the effluent met discharge standards.

Mr. Delahaut is very pleased with the new process. "I'm a small business owner and did not have much capital for buying cleaning equipment. I was able to purchase a low-cost spray cabinet that cleans very effectively. The water process has the added benefit of being safer for the workers and my neighbors."

Hydro-Aire, a division of Crane located in Burbank, has 572 employees. The aerospace firm manufactures braking systems, pumps and air locking devices and serves as a subcontractor for Boeing. Hydro-Aire also does repair of the pumps used in military and commercial aircraft like the C-130 transport and the C-17.

Hydro-Aire assembles printed circuit (PC) boards for commercial and military use. In the board assembly process, flux is applied to the boards to facilitate solder flow, to effect heat transfer and to prevent oxidation of the boards. The components are soldered to the boards and then the boards must be cleaned to remove the remaining flux and any other contaminants that might remain on the boards. Hydro-Aire had two different cleaning operations for cleaning the boards. One of these, where the boards were targeted for commercial applications, relied on water-soluble flux. The flux was cleaned from the boards with plain deionized water. The other operation, where the boards were used in the military sector, used a rosin flux and 1,1,1-trichloroethane (TCA) for removing the contaminants from the boards.

IRTA began work with Hydro-Aire last year and suggested that the firm begin exploring alternatives. Production of TCA was banned in 1996 but a supply still remained. The price of TCA was very high, however, at about $8 per pound. Hydro-Aire would have to find an alternative before the supply of TCA ran out and they could reduce their costs if they converted to an alternative cleaning system. In addition, the Halogenated Solvents Cleaning NESHAP became effective on existing facilities in December, 1997 and Hydro-Aire would have to comply with the equipment and recordkeeping requirements. Converting to an alternative would allow the firm to avoid being subject to the NESHAP.

There are two water-based processes available for cleaning PC boards. First, if water soluble flux is used on the boards, deionized water with no additives can be used for cleaning. Second, if rosin flux is used on the boards, a water-based cleaner with saponifier additives must be used to remove the flux. Hydro-Aire was already using water soluble flux in the commercial operation but had existing military specifications that required the use of rosin flux in the military operation. All new military contracts allow the use of water soluble flux but existing contracts often still require rosin flux.

After considerable testing and analysis, Hydro-Aire purchased a large batch dishwasher from Aqueous Technologies and settled on a water-based cleaner with no solvent additives for the cleaning process. The firm will continue to use rosin flux but has the option of converting the operation to water soluble flux later when the contract expires. The cleaning unit does not require an air district permit because the cleaner has less than two percent VOC.

For disposing of the spent cleaner, Hydro-Aire opted to purchase an evaporator. Testing is underway to determine if the spent cleaner is hazardous waste. In that event, a tiered permit from Cal-EPA's Department of Toxic Substances Control would be required. No air district permit is required because the spent cleaner contains no VOCs or toxics.

Tommy Jennings, the Environmental Manager at Hydro-Aire, is in the process of implementing the new system. "We now have a process which is good for the environment and the workers and it is giving us good results on the boards," says Mr. Jennings. "The operation is not covered by the NESHAP and our costs are down because of the high price of TCA."

In 1997, IRTA began work on a project sponsored by the South Coast Air Quality Management District (SCAQMD) to work with companies on further reducing VOC emissions in cleaning operations. IRTA began looking for firms in Southern California to participate in the project. A firm that expressed interest in the project was Astro Pak, one of the nation's leading precision cleaning contractors specializing in the cleaning of high purity gas and fluid systems. The company provides precision cleaning services for pipes, tubing, components, tanks, hoses and fittings for almost every industry, including aerospace, military, pharmaceutical, microelectronics and semiconductor.

It is commonly thought that solvent cleaning agents are required for precision cleaning operations. Many people in the industry hold the view that water-based cleaners are suitable for general metal cleaning but that they cannot really be effective for precision cleaning.

When IRTA began work with Astro Pak, the firm was using 1,1,1-trichloroethane in a vapor degreasing process to preclean most of the hardware they process. The Halogenated Solvent Cleaning NESHAP was scheduled to become effective for existing processes on December 2, 1997. Astro Pak decided that the best long-term option would be to convert to a water-based cleaning process. Over the short-term, while the company worked with IRTA to complete the conversion, Astro Pak converted to n-propyl bromide (NPB). The firm did not want to use nPB permanently for environmental reasons. The chemical is classified as a VOC and it contributes to stratospheric ozone depletion.

Astro Pak's precleaning operation needs to be flexible enough to remove a variety of contaminants from many different substrates. The soils encountered by Astro Pak include hydrocarbon grease, oils, krytox, other halocarbon greases and general dirt and particulates. The company routinely cleans stainless steel, brass, aluminum and carbon steel as well as Buna rubber, Viton and Teflon gaskets.

IRTA arranged for preliminary laboratory testing of a set of Astro Pak's parts to identify a few water-based cleaners that would be suitable for removing the soils. Because of the complex configurations of many of Astro Pak's parts, IRTA and Astro Pak agreed that an ultrasonic cleaning system would be necessary. The cleaner that performed best in the laboratory setting was further tested in a small ultrasonic cleaning unit at the Astro Pak site. The hardware cleaned with the water-based cleaner, which is made by W.R. Grace, passed the nonvolatile residue (NVR) analysis and particulate count tests.

After carefully considering their cleaning needs, Astro Pak purchased a three-bath cleaning system built by CAE Blackstone. The first tank, which holds about 750 gallons, is the ultrasonic wash bath. The second two tanks, which also hold 750 gallons, are rinse tanks. All three tanks are fitted with spray wands so the parts can be flushed. The system includes filtration and oil removal capability.

The new system was installed in October, 1998. The cleaner, which rejects oil, lasted about four months before it needed to be changed out. Says Brent Ekstrand, Astro Pak's cleaning expert, "we cleaned one particularly dirty part in the system. The bath life is likely to be much longer the next time."

Mr. Ekstrand conducted a set of tests to determine if the new water-based cleaning system was cleaning as well as the nPB. He measured the NVR on the same parts after cleaning with nPB and after cleaning with the water-based cleaner. The lower the NVR, the cleaner the part. Out of 25 different pieces of hardware tested, 21 showed a lower NVR when cleaned with the water-based system. Three of the parts had the same NVR when cleaned with the NPB and the water-based system. Only one part had a lower NVR when cleaned with NPB. "The results are statistically significant," says Mr. Ekstrand. "Our parts are cleaner with the new water-based system."

Astro Pak is saving nearly $1,800 per month by converting to the water cleaning system in spite of the cost of the new equipment. According to Ken Verheyen, Astro Pak's owner and President, "we're saving money and we're cleaning more effectively with water than with solvent. As an added benefit, we're avoiding the liability a solvent always carries. Astro Pak wants to be good to the workers and the environment and the water cleaning system helps us do that."