Solvent Mileage. Government Involvement. Introduction. Increased Profit. Calculating Solvent Mileage. When to Calculate

Transcription

1 Vol. 12, No. 1 0 INTERNATIONAL FABRICARE INSTITUTE March 1988 Solvent Mileage Introduction When was the last time you accurately checked your solvent mileage? Having a general idea of how many gallons of solvent purchased per month or per year is not sufficient to give an indication of solvent mileage. Keeping accurate records requires taking time to weigh, record, and total the load sizes and the solvent and detergent added to the machine. There are three major benefits from increased solvent mileage: increased profit as a result of lower solvent purchases decreased atmospheric emissions, and fewer solvent vapors present in the plant. Increased Profit Anyone asking why increase solvent mileage? should understand that low solvent mileage and high solvent consumption play a major role in the plant s profit picture. When was the last time you calculated how much low solvent mileage costs you in a month or a year? As an example, take several average fisures for a perchloroethylene plant: Weekly cleaning volume 3000 lbs. Current solvent mileage 6300 lbs/drum (52 gals) Price of perc $4.50 gal. Potential solvent mileage 11,000 lbddrum With a mileage figure of 6,300 pounds per drum and 3,000 pounds of garments a week, 25 drums of solvent are used in a year or approximately 2 drums a month, If the solvent mileage is increased to 11,000 pounds per drum, only 14 drums a year are used. The difference is 11 drums of solvent a year, or 572 gallons of perc. At $4.50 a gallon, that totals $2, a year. Tables 1 and 2 enable you to quickly obtain the costs of poor solvent mileage for perchloroethylene and petroleum solvents. Government Involvement Solvent recovery isn t the only reason for being concerned about solvent mileage. Two government agencies regulate solvent emissions. The Environmental Protection Agency (EPA) regulates how much solvent vapor can be emitted into the atmosphere by a plant. Solvent vapors released inside the plant are the concern of the Occupational Safety and Health Administration (OSHA). Many of the factors contributing to low solvent mileage result in a higher level of vapors in the cleaning area (for example, liquid leaks, worn door gaskets and poor tumbler seals). Obtaining good solvent mileage almost always leads to a significant reduction in solvent vapors in the cleaning area. Although government regulation is often resented, OSHA and EPA have benefited the drycleaning industry in two ways. By enforcing good housekeeping, proper operating procedures, and testing control equipment, drycleaners have increased solvent mileage. For instance, controlling leaks and replacing worn or damaged gaskets and seals can have a dramatic effect on solvent consumption. Reducing vapor levels in the plant has improved working conditions and has led to increased worker productivity by helping plants maintain a more pleasant working atmosphere for employees and customers. Calculating Solvent Mileage When to Calculate When checking solvent mileage, be aware that the mileage figure only represents the effect of the conditions of that particular season. The difference in incoming cooling water temperature from winter to summer can drastically affect solvent recovery and mileage. If calculating solvent mileage is not a continuous procedure in your plant, you should spot check solvent consumption at least twice a year in different seasons for optimum comparison. INTERNATIONAL FABRICARE INSTITUTE

2 Monitoring solvent mileage on a continual basis has its advantages. By keeping accurate records all year, you will know if your mileage is in the expected range. When mileage starts to deviate, you are immediately alerted to improper procedures or minor problems that could become costly. What Records to Keep 5) In order to get an accurate solvent consumption figure, you must maintain reliable records over a set time period. You will need to keep records of the following items: the weight of every load of garments cleaned, the amount of new solvent added, and the amount of detergent used. If the solvent returned from separate recovery tumblers, carbon adsorbers, or from reclaiming cartridges or filter muck is not added back into the operating system, you must also keep track of these amounts. The Calculation 1) Run your machine a few minutes to make sure the filters are full of solvent. 2) 3) 4) 6) 7) Let the solvent in the wheel and pipes drain back into the base tank. Add solvent to bring the solvent level up to a premarked point or until the tank is full. Keep reliable records of the pounds of garments cleaned over a set period of time (for example 3 weeks). At the end of this time period, repeat the process outlined in steps 1-3 and refill the base tank to the same level as in step 3, keeping track of the amount of solvent added. To determine the number of gallons of solvent consumed, add the amount of solvent needed to refill the tank plus the amount of new solvent and detergent added to the machine since the beginning of the evaluation period. Subtract from this amount any solvent (such as from recovery tumblers, vapor adsorbers or dip tanks for sizing, etc.) that was not returned to the base tank. To calculate your solvent mileage, use the following formula: Solvent Mileage = total pounds of clothes cleaned gallons of solvent consumed Table 2. Savings Per Year with Increasing Petroleum Solvent Mileage to 2,250 lbs/drum Cleaning Volume Present Solvent Mileage (lbs cleaned/52 gal drum)* lbslweek 500 (10) 1,000 (20) 1,500 (30) 2,000 (40) 1,000 2,000 3,000 4,000 5,000 7,500 10,000 15,000 25,000 $ 5, , , , , , , , , $ 1, , , , , , , , , $ , , , , , , , , $ , , , , * Numbers in parentheses are approximate pounds cleaned/gallon solvent. Savings based on petroleum solvent cost of $1.2O/gallon and possible mileage of 2,250 lbs/drum. ~ 2 FOCUS ON DRYCLEANING 3/88

3 Multiply this amount by 52 to obtain the number of pounds cleaned per drum. Use the solvent mileage work sheet to easily calculate your solvent mileage. 'hbls 8. Drycleaning Solvent P$ctors w*t.rrolxune 80hient [IWgel) in gal Ibsldrum oroletthylena sa 700 on a2 $40 Where the Losses Occur Liquid Solvent Leaks Liquid solvent leaks are frequently found in drycleaning plants. In perc plants, these types of leaks occur almost exclusively in three areas: Washer door, Still or cooker door, Pipe unions. In petroleum plants, bearings, washer doors, pipe unions, and 4-way filter valves account for the majority of leaks. As seen from Table 4 considerable amounts of solvent can be lost each month from a single drip. You can see that even one drip every five seconds which continues for eight hours a day will amount to 12 gallons of solvent lost per month. The solvent loss from a liquid leak such as this is almost completely nonrecoverable. Usually, the drop of solvent hits a large surface area-such as the floor or part of the drycleaning equipment-and then evaporates rapidly. Good housekeeping cannot be overemphasized in order to prevent this type of solvent loss. It is difficult to locate liquid solvent leaks when the machine is dirty. Many times leaks occurring from the rear of the drycleaning equipment go completely unnoticed because the drops fall into an oily lint buildup on the back of the machine. Leaks from underground storage tanks in petroleum plants may be a more prevalent condition than has been supposed. There have been indications of leaks in underground tanks of 30 to 40 gallons per day, but the actual losses could be higher than this. A crack in an underground tank will result in a solvent loss if the soil around the tank is dry and porous. If the soil is wet, you're likely to find that hydrostatic pressure will result in water being forced into the tank. If you find that you have far more moisture in your solvent than you've added, this could be the cause. Solvent Vapor Leaks Solvent vapor leaks are another area of concern. Five consistent areas contribute to vapor leaks: Washer door, Button trap, Still or cooker doors, Recovery tumbler doors, Tumbler or lint trap door. Also contributing to solvent losses in plants are open buckets, button traps, and water repellent containers. Solvent will evaporate from any open container into the air. Dryers and Recovery Tumblers The loss of solvent during the drying operation is governed mainly by extraction efficiency. Extraction efficiency depends on the following factors: The type and diameter of extractor, Revolutions per minute (RPM), Extraction time, Load composition. In petroleum plants, the solvent retained in the load after extraction will be completely lost during drying if no recovery tumbler is availble. Retention after extraction can vary widely but should fall in the range between 8 to 21 percent by weight [from 8-12 pounds solvent per 100 pounds dry garment weight). In perchloroethylene plants, high solvent retention after extraction is not as damaging to solvent mileage because a recovery tumbler is used. However, when retentions reach 50 percent, if even one factor involved in proper recovery is marginal, higher than normal solvent losses will occur. According to our past plant studies, a typical mixed load should retain approximately 25 percent after proper extraction. Improper operation of recovery tumblers accounts for more lost solvent than any other source. The single most important factor in obtaining good recovery efficiency is allowing the load to dry long enough (excluding aeration time). The normal minimum drying time with all other operating factors correct is minutes. With a bone-dry synthetic load (no moisture added in cleaning) this might drop to minutes. A drying time of minutes is a starting point; times increase or decrease depending on the load type. If the weights of two different loads are the same, then the load with thicker and more absorbent fabrics will dry the slowest. For example, while a 50 pound load of silks might dry in the minimum time, a load of wools with the same weight would require 5- ~~ 3/88 FOCUS ON DRYCLEANING 3

4 10 minutes additional drying time. The three important factors in achieving good recovery efficiency include: the minimum drying time; added time if any moisture is used in cleaning: and added time for increasing fabric thickness. Calculating Reclamation Efficiency and Extraction Efficiency It is wise to have a recent calculation of recovery and extraction rates close at hand. By keeping continuous current records, you'll know that any difference in the efficiency figures is an indicator that something needs to be changed. If possible, remove the water separator so you can catch the solvent and water coming directly from the condenser. If this isn't practical, make sure the separator is primed before starting. You will also need an accurate scale and a container to catch the water andperc. For best results you should repeat the following procedure several times and to obtain an average figure. To obtain your reclamation and extraction efficiencies, use the following procedure Load the machine and run a normal cleaning and extraction cycle. [When repeating this process always use the same weight and type of load to get consistent results.) Immediately after the extraction cycle, remove the wet load and weigh it carefully. Record this as the wet weight. When handling wet loads, you should wear an organic vapor, and solvent resistant gloves. At the recovery tumbler, disconnect both the water and solvent lines coming from the water separator. Reroute them into two separate, preweighed containers. Put the load in the tumbler and dry normally, collecting perc and water in the containers. After the aeration cycle, weigh each container separately. Subtract the weight of the containers to obtain the weight of perc recovered, and the weight of water recovered. Weigh the dry load and record this as the final dry weight. You can now calculate the reclamation efficiency with the following formula: % Reclamation - perc recovered x 100 efficiency wet -water - final dry weight recovered weight We recommend that you determine your reclamation efficiency once in warm weather and once in cold weather. You can now calculate extraction efficiency with the following formula: % Extraction = wet weight - final dry weight Efficiency final dry weight Use the worksheet to calculate your plant's reclamation efficiency and percent extraction efficiency. Additional Factors -~ Other factors that affect solvent reclamation follow in their approximate order of importance: Reruns and Underloading. Every time you run a load of approximately five pounds or less in a 50 pound capacity tumbler, solvent recovery is O! This will hold roughly proportional for other size tumblers, i.e., four pounds in a 40 pound tumbler, etc. Inside the tumbler, the amount of perc vapors in each cubic foot of air can only be dropped to a certain level, depending on the condenser temperature. These are the vapors that exhaust during aeration. In a 50 pound tumbler, the perc left in a five pound load after extraction is only enough to bring the vapor concentration up to the point at which the condenser would first begin to recover solvent. Also, higher temperatures are reached in a short time with very small loads. In this situation, heat-sensitive fibers may be affected when run by themselves, but not when run with a full load. Underloading a tumbler by not more than 20 percent can speed recovery somewhat: underloading by much more can affect recovery efficiency because of the same mechanism mentioned previously. Drying Temperatures. This factor is very closely related to water temperatures. A change in the drying or water temperature will significantly change the acceptable limits of the other. The drying temperature of most importance is that of the air after it leaves the cylinder but before it reaches the condenser. This is the stack temperature and should be in the range of 135'-145' F. This temperature range is safe for almost all modernday fibers. If the drying temperature is dropped ZOO F, approximately five minutes more must be added to the drying cycle to get the same recovery efficiency. (Note: if you change two or more factors at one time, you have to add the extra time for each one to your basic drying time!) Water temperatures. The water temperature of concern is the water leaving the condenser. The cooler this water is-down to 55O-60' F-the more efficient is your solvent recovery. This temperature range is a rough practical lower limit: when the condenser temperature is below this, the steam coils may not be able to bring the air temperature back up to an acceptable level for drying. If your condenser water outlet temperature is in the range of 100 -lloo F, approximately five minutes more must be added to the drying cycle. Leaking Aerate Dampers. There are two aerate dampers on your tumbler, one for air inlet and one 4 FOCUS ON DRYCLEANING 3/88

5 for exhaust. During the reclamation cycle, when the garments are being dried, the dampers should be tightly closed to prevent any solvent-laden air from escaping. A problem arises when a damper does not shut tightly during the recovery cycle and a leak occurs. IF1 recently conducted tests to measure how much solvent is lost through damper leaks of various sizes. A battery powered sampling pump and a gas chromatograph were used for analysis of the air which leaked at the damper location. These test runs were conducted in a 40 pound transfer machine and the findings are listed in Table 5. lation losses, this loss of solvent can be substantial. In all three solvents, there are two factors which influence solvent losses in filters. The first factor is cartridge life. Longer cartridge life will increase solvent mileage and shorter cartridge life decreases solvent mileage. The time allowed for draining andlor drying of the cartridge before discarding affects solvent loss considerably. We recommend draining the cartridges at least overnight and preferably over a weekend. This applies to cartridges in all three solvents. Water Separators In perc plants, water separators can be another source cd solvent loss. Lint in a tumbler separator or rust and corrosion inside the still, cooker, or adsorber separators can cause a problem by blocking the solvent drain line or the vent lines that equalize the pressure in the separator. When this happens, the solvent level in the separator may rise high enough for solvent to flow out the water drain line. The smallest leak (.02 to.16 inch] represents the size of leak that could be caused by a worn damper or lint build-up. The larger leaks are representative of the gaps that could occur due to mechanical linkage failure. Losses From Filters, Stills, and Cookers. The solvent lost in residues from filters, stills, and cookers can affect solvent mileage tremendously in all three solvents: petroleum, perchloroethylene, and fluorocarbon. In petroleum plants, solvent losses in residues are almost always the second highest source of loss. Petroleum Filter Residues. According to the results of the membership survey run by the IF1 Research Department in the mid-seventies, approximately 81 percent of petroleum plants use either rigid tube, screen, or bag type filters. In petroleum plants, even after draining and airdrying while still inside the filter, the average loss of solvent when the residue is removed is going to be about 14.5 gallons per 1000 pounds cleaned. Put in a different way, every 100 pounds of sludge removed from the filter is going to contain over ten gallons of solvent. IF1 is witnessing a change to regenerative and cartridge filtration systems. Regenerative filters can reduce solvent losses approximately by half, as less residue is produced for every 1000 pounds of cleaning. A change to cartridge filters would reduce the solvent losses per 1000 pounds of cleaning to less than one-tenth of the amount lost with screen, rigid tube, or bag filters, and labor costs related to filtration would also drop. Cartridge Filters General losses of solvent in the cartridges are relatively low; however, when combined with the distil- Vapor Adsorbers When properly used, a vapor adsorber is one of the best tools available to the perc cleaner for increasing solvent mileage-but it was never designed to be used as a crutch when maintenance and efficiency of other equipment is substandard. An adsorber is very efficient when properly operated. Over 120,000 parts per million can be put into it from a tumbler on aerate, but the adsorber exhaust will be 15 parts per million or less-that s efficiency! If you overload the adsorber by not stripping often enough or by not drying the bed out after stripping, your adsorber becomes a tremendous exhaust fan releasing as much as 1,000 parts per million of solvent. At this point, until you can properly strip the adsorber, it would be better economically to turn the adsorber off and duct your washer and tumbler outside. Why? The standard single-bed models in our industry have a fan with a capacity of 600 cubic feet per minute (cfm); 1,000 parts per million in the exhaust being pushed out at 600 cubic feet per minute means a loss of over one gallon every hour. Effect of Operational Variables on Solvent Mileage From the various surveys of manufacturers/ distributors, IF1 surveys, internal IF1 research, and the Federal Government (EPA) studies, there is a solid basis to report an average nationwide solvent mileage. This publication attempts to present a realistic picture of how the solvent mileage varies under conditions of different operations and equipment. 3/88 FOCUS ON DRYCLEANING 5

6 ~ INTERNATIONAL FABRICARE INSTITUTE However, because of the obvious variations between plant operations, we will refer to an average solvent mileage upon which our calculations and results are based. Also, these calculations are based on the assumption that all equipment is properly used and maintained. The basis for the estimates in Table 6 is studies conducted at IFI. The brief summary of estimated loss follows: Evaporated at the Washer. This figure is based on measurements supplied by a drycleaning machine manufacturer. The measurements were taken of the lost solvent vapor that was collected in vapor adsorber from the washer of 55 pounds and 85 pounds capacity. Evaporated at the Dryer. An average load of garments should contain approximately 25 percent after proper extraction. With good recovery efficiency, about 12 percent of solvent is lost during aeration. This is about 30 pounds of perc per 1,000 pounds per load, or 2.2 gallons per 1,000 pounds. Exhausted from Adsorber. A properly operated adsorber will have a small quantity of perc vapors in its exhaust (15 parts per million is an average). Since the loss depends on operating time of the adsorber, rather than on pounds of clothes cleaned, this figure is an approximation, based on 1,000 pounds cleaned for every 12 hours of plant operation. Retained in Filter Muck. The quantities of filter powder, carbon, detergent, and soil [all nonrecoverable residues) used per 1,000 pounds of cleaning were derived from several IF1 Fellowship bulletins. For a rigid tube filter with no cooker, about 60.2 pounds residue per 1,000 pounds cleaning, which retains about 140 pounds (233%) of perc after draining only. After cooking, the perc retention drops to about 17 pounds when a steam- or air-sweep is properly used. Finally, with regenerative filters we found about 32.9 pounds of residue per 1,000 pounds cleaning, retaining 9.2 pounds after cooking. Retained in Paper Cartridge Filters. In previous IF1 Fellowships 21 cartridges were used in cleaning 14,493 pounds of clothes. The average solvent retention of the cartridges after draining overnight gave a loss of 1.3 gallons per 1,000 pounds cleaned. By putting the cartridges in a drying cabinet vented to an adsorber, this loss could be reduced to 0.9 gallons per 1,000 pounds cleaning. Retained in Still Residue. According to IF1 tests performed in the early ~ O S, there was a 2.3 percent loss of perchloroethylene in distillation. Based on minimum recommendations of 5 gallons distillation per 100 pounds cleaning, this is equivalent to a loss of 1.2 gallons of perc per 1,000 pounds. 6 FOCUS ON DRYCLEANING 3/88

7 Leaks. These are miscellaneous evaporative losses that occur from evaporation during extraction, during transfer, and small drips. This is an estimate as it is almost impossible to measure these types of losses directly. On the basis of our estimated figures, let s look at the effect of various plant operations and how they affect the solvent mileage. First, there is a difference between a plant that has a vapor adsorber and the one that has not. Both plants use cartridge filtration, and the extraction efficiency is in the average range of 25 percent extraction retention. As you can see, the difference is 4,502 pounds per drum, or 36.4 percent better mileage with a properly operated vapor adsorber. Now let s look at the plant that uses regenerative filters and compare mileage with a plant without a vapor adsorber, assuming extraction efficiency is the same. The difference is 10,834 pounds or 50 percent more cleaned with the vapor adsorber. This difference is even greater if we compare the above mileage with cartridge filtration. This is because in cartridge filtration there are combined losses of solvent from cartridges and still residues. Therefore, in cartridge filtration, the reduction of perchloroethylene solvent is essential. Let s now look at the extraction efficiency factor. But first, let us explain how we arrive at pounds of loss solvent in 100 pounds cleaned load under the different extraction and recovery efficiencies. Example: Cleaning a 100 pound load at 60 percent extraction efficiency will contain 60 pounds of perchloroethylene. If the recovery efficiency is 80 percent (recovered during drying cycle], 20 percent of solvent is lost from the original 60 pounds of perc remaining in the garments after extraction. Therefore, 20 percent of 60 pounds is 12 pounds of perchloroethylene lost. From the examples in Table 7 you can see the relationship between different retentions and recovery efficiencies, and their effect on mileage of solvent, The example in Plant D shows the effect of an ideal load consisting of wools. Wools have the lowest solvent retention and cotton the highest. During IF1 in-plant studies, the average solvent retention was found to be about 25 percent after proper extraction. With the advent of new technology used in designing drycleaning machines, the perc solvent mileages have increased substantially. The examples in Table 7B are comparing solvent mileage of a closed machine on the basis of IF1 Fellowship evaluation results, and the realistic increases of solvent mileage by stripping cartridges and the use of a new method of reducing perc content in still residue. In this table we also compare the conventional drycleaning machine with cartridge filtration with and without adsorber with emphasis placed on the perc reduction from still residues and cartridges. If you compare solvent mileage of Machines C and D with comparable machines where the attempt was made to reduce perc content in cartridges and still residues, you ll see that it can result in monetary savings. Petroleum In most petroleum plants, the largest cause of solvent loss is the drying tumbler. However, about a decade ago, the Hoyt company designed recovery tumblers for petroleum solvents. This equipment is available for petroleum plants now. Taking into consideration that the majority of petroleum cleaners still use conventional tumblers where the solvent is lost, and considering some of the recent studies in petroleum plants, the estimate for current average mileage of 1,200 pounds per drum (231 pounds garments per gal.] is derived. With good extraction efficiency achievable mileages are 1,600-1,700 pounds per drum with a regenerative filter and 3/88 FOCUS ON DRYCLEANING 7

8 Table 7B. Loss of Solvent/ 1,000 Ibs. Cleaned a) Washer Loss b) Tumbler Loss Machine A No Vent Refrigeration, Cool Condenser, Condenser, No Adsorber 0 gal. 0 gal. Machine B Same as Machine A Except Cartridges Stripped of Perc by about 70% 0 gal. 0 gal. Machine C Regular Machine Without Vapor Adsorber, Cartridge and Still Residue Reduced by gal. 2.2 gal. Machine D Same as Machine C Except With Adsorber 0 gal. 0 gal. c) Exhausted from Vapor Adsorber d) Retained in Cartridge- Drained e) Retained in Still Residue 0 gal. 1.3 gal. 1.2 gal. 0 gal gal. 1.2 gal. 0 gal. 0.6 gal. 0.6 gal. 0.2 gal. 0.6 gal. 0.6 gal. 2.6 gal/ 1,000 cleaned = 20,000 lbs/drum 1.65 gal/ 1,000 cleaned = 31,515 lbs/drum 5.3 gal/ 1,000 cleaned = 9,811 lbs/drum 2.9 gall 1,000 cleaned = 17,930 lbs/drum 2,200-2,300 pounds per drum with cartridge filtration. In perchloroethylene plants, operational conditions-rather than the types of equipment-are usually the limiting factor for solvent mileage; in petroleum plants the reverse is most often true. For example, the second highest source of solvent loss in petroleum plants is in filter residues and at the present time the only way to significantly change losses from this source is to change the type of filter. Table 8 contains estimates of the amount of petroleum solvent lost at different stages of the drycleaning process. These are average losses assuming that all equipment is properly used and maintained. While losses from an underground storage tank cannot be considered normal in that it is not a part of the drycleaning process, this may be a more prevalent condition that has been supposed and therefore it is included in this table. - 8 FOCUS ON DRYCLEANING 3/88

9 A brief summary of these losses is given below, Evaporated at Dryer. As reported in IF1 bulletins and from the results of recent IF1 Research studies, ranges of solvent retention after extraction (on a weight basis) were found to be: 1) separate extractors:: , 6) front loading washer/ extractor: 10-18%, and c) side loading washer/ extractor: 1341%. Retained in Filter Muck. The quantities of filter powder, carbon, and soil used per 1,000 pounds of cleaning were derived from previous Fellowship testing. For a screen or rigid tube filter, there was approximately 47.6 pounds of residue per 1,000 pounds of cleaning. This residue retains about 200 percent of petroleum solvent (by weight) after draining for 24 hours. With regenerative filters, about 23.8 pounds of residue per 1,000 pounds of cleaning was found, also at 200 percent retention after draining (if this drained residue is properly air-dried 25-50% of the solvent remaining may possibly be recovered). Retained in Cartridge Filters. A typical cartridge retains about 0.85 gallons of solvent after thorough draining. The loss per 1,000 pounds of cleaning was calculated on the basis of a cartridge life of 700 pounds garments. Retained in Still Residue. A solvent loss of 3.0 percent is typical in vacuum distillation. Based on IF1 minimum recommendations of five gallons distillation per 100 pounds of cleaning, this is equivalent to a loss of 1.5 gallons of petroleum solvent per 1,000 pounds garments. Miscellaneous Evaporative Losses. These are losses that occur from evaporation during extraction and transfer, small drips, and so on. This is an estimate as it is almost impossible to directly measure these types of losses. Leaks in Underground Tanks. A leak runs 7 days a week, regardless of the poundage cleaned. This calculation is based on a plant doing 3,000 pounds per week. With this cleaning volume the loss amounts to about 70 gallons per 1,000 pounds cleaned. The following examples illustrate the major areas of petroleum solvent loss. What is significant here is that by changing filtration systems, losses in residues are cut tremendously. This is not the solution for everyone, but it is an alternative that may be considered. Compare Plant I and 111, which are identical except for the filtration system; mileage has increased by 901 pounds per drum by using a cartridge filter. When translated to more practical terms, a plant doing 3,000 pounds per week with a screen, rigid tube, or filter and pay $1.20 a gallon for petroleum solvent, would save $2,490 a year in solvent costs after converting to cartridges. Given the same conditions, changing from a screen, rigid, or bag filter to a regenerative would sve $1,274 annually, and from regenerative to cartridge $1,216 annually. If a change is made to cartridge filtration, the benefits would be direct solvent savings and reduced labor costs. On the negative side are the initial installation/changeover costs and the higher direct filtration costs per pound of load. Following is an actual example of a plant which had leaking underground tanks. Example 4: Plant doing 4,800 lbs/wk.; using 415 gal/wk. Underground leak of 207 gal/wk. Rigid tube filter; still: side-loading washedextractor. Measured extraction efficiency: 17.2% retention. At the rate of 207 gallons a week, the leak from their underground tank is costing this plant about $13,000 annually. Table 9 can be used to determine either your current solvent consumption (if you know your present mileage) or your potential mileage. 3/88 FOCUS ON DRYCLEANING 9

10 Mileage without Leak a) Evaporated at dryer 127 gal./wk. (17.2% retention) b) Loss in muck (tube filter) 69 gal./wk. c) Still residue loss 7 gal./wk. d) Misc. loss 5 gal./wk. Weekly Total Loss: Equivalent Total Loss: Solvent Mileage: Fluorocarbon Plants 208 gal./wk ga1./1000 Ibs. 1,200 lbs/drum Since the drycleaning machines that use fluorocarbon solvent are built virtually air tight, leak free, ventless, and are equipped with refrigeration-type condensers, it would seem that achieving high mileage wouldn't be a problem. Then why is there such a wide fluctuation in fluorocarbon mileage? The average mileage of fluorocarbon solvents is between 8,000 and 20,000 pounds cleaned per drum. The main problem lies in the operation of refrigeration condensing system. Many plants that experience less efficient operation of refrigeration systems are not aware of it until they notice that they are spending more solvent on cleaning. As an exapmle, during the IF1 in-house tests, the average mileage was 14,800 pounds cleaned per drum of solvent. When the refrigeration system was operating properly, this mileage jumped to 22,000 pounds cleaned per drum. When the refrigeration system wasn't operating properly, the mileage dropped to 9,600 pounds per drum. Mileage with Leak a) Weekly Total Loss (above) 208 gal./wk. b) Loss from Underground Leak 207 gal./wk. Weekly Total Loss: Equivalent Total Loss: Actual Solvent Mileage: 415 gal./wk ga1./1000 lbs. 601 lbs/drum From IF1 Valclene" studies, solvent losses from the still residue were 0.2 gallons of solvent loss per 1,000 pounds cleaned. The tests distilled 129 gallons of solvent per each 1,000 pounds cleaned and generated 1.2 gallons of still residue. The solvent loss during distillation (considering 0.2 gallons solvent loss per 1.2 gallons of generated still residue per 1,000 pounds cleaned), was 16.7% on a volume per volume basis. After the recommended procedure for drying cartridges in the cleaning wheel, there was not a significant quantity of solvent remaining in the cartridges. Still Residue and Filter Loss According to IF1 Still Residue Analysis data, the fluorocarbon solvent loss was 1.41 gallons per 100 pounds of still residue. 10 FOCUS ON DRYCLEANING 3/88

11 Figure 1. How to Improve Solvent Mileage When you are aware of where solvent losses occur, you must act to measure and minimize them. Proper Operation of Stills and Cookers In petroleum distillation, there are a few factors that are easily checked to minimize losses. The average loss of petroleum solvent in vacuum distillation should be three percent of the total solvent distilled. That is, for every 100 gallons of solvent distilled, about three gallons will be lost in the residue. This figure is already corrected for the nonvolatile residue (NVR) present. The total quality of Eesidue (including solvent) removed from the still is going to depend on the percent NVR present in your solvent. If you have one percent NVR in your solvent, the total residue removed from the still after distilling 100 gallons would be one gallon (NVR) plus three gallons (solvent loss) = four gallons total. Of the ways to prevent losses in vacuum distillation, the most important is to insure that there is not an excessive accumulation of nonvolatile residue in the still itself. This will raise the boiling point of the solvent in the still and make it more difficult to distill the solvent. As an example, in a test run at IFI, distillation had stopped after a little over two hours with the system being operated at 27 inches of vacuum and 27 pounds steam pressure. Even though there was no solvent coming over, the residue contained in the still was about 13 percent by weight NVR and 87 percent solvent. Between 23 and 24 gallons of solvent would have been lost if the residue had been dumped. What's needed at this point is to open the bypass around the reducing valve and boil down the still using full steam pressure. In perchloroethylene plants, either a normal still unit or a combination still/muck cooker may be used. A large advantage that perchloroethylene plants have over petroleum plants in increasing solvent mileage is the ability to remove solvent from filter residue by using a cooking operation. In distillation of perchloroethylene, there are several areas of operation to monitor in order to minimize solvent losses. The first factor is to make sure that distillation is complete before shutting down the still and discarding the residue. Solvent flow from the still water separator back to the cleaning machine should be reduced to a trickle before stopping distillation-as a rough guideline, no more than one cup to one pint of water should accumulate in 15 minutes. For proper distillation, steam pressure should be in the range of pounds per square inch. Higher pressures mean higher temperatures and the solvent will distill over a much faster rate. This can cause a loss of solvent, as the still condenser may not be able to sufficiently cool the perc and water mixture coming over. When this happens, the result is perc carried over into the machine base tank. Another problem with high steam pressures is that undesired impurities may distill over the solvent. If you find that distillation won't start in the normal pressure range of pounds per square inch, check the type of steam trap that you have: the most common ones are bucket traps and impulse traps. (Figures 1 and 2.) Impulse traps work well at the medium steam pressure that should be used for distillation, while many bucket traps are just beginning to operate at pounds per square inch. 'The other part of insuring that you have efficient separation and recovery of solvent during distillation lies with the cooling water temperatures. The temperature of the water coming out from the still condenser shouldn't exceed 110" F. Preferably, this should be checked with a thermometer-but if you Figure 2. ~~ 3/88 FOCUS ON DRYCLEANING 11

12 can't comfortably hold your hand in the water corning out, it's too hot. Most people won't willingly tolerate a water temperature above 115"-120 F. Of course, the critical temperature is that of the solvent/water mixture in the separator. Check the temperature of the solvent in the drain line going back to the base tank. It should be 80" F, but never over F. If it is higher, increase the water flow through your still condenser. If you continue to have problems with return solvent temperature, after-coolers are available for the separator. If you already have an after-cooler, the water temperature should be about 75" F. One last area to check on your still is the vent pipe. Because pressure is built up inside the still during distillation, an atmospheric vent is necessary. Although the pressures involved are usually very low, some saturated solvent vapors will be forced out of the vent. Since the vapor/air mixture is at a relatively high temperature and is saturated, the concentration of perc could be 240,000 parts per million-or even three times higher than that! If you have a vapor adsorber, the best procedure is to run a stand pipe up from the vent and connect the end of the pipe to your adsorber ducting with a piece of flexible tubing. For safety reasons, if you don't have a vapor adsorber, the vent should be piped outside the building to a point above the roofline. Solvent vapors will usually condense on the cool walls of the pipe or duct within six feet above the still, after which they drain back. To minimize solvent losses, many stills and cookers are equipped with a steam- or air-sweep. If this operation is not done properly, the final solvent recovered can be loaded with odor-causing contaminants. Steam or air should enter the kettle under the low pressure (usually 5-7 pounds per square inch) for only a short time (3-5 minutes). This procedure takes place at the end of the boil or cool down while the steam is still passing through the steam coil, thus distillation continues. Some operators find it advantageous to use a by-pass around the reducing valve to increase the steam pressure to about 80 pounds per square inch during the boil down. However, if any foaming is observed, the steam pressure should be lowered until it stops. Measuring Perc Content in Still and Cooker Residues Periodic monitoring of the perc content in still and cooker residues can be done in your plant and is an important indicator for evaluating solvent mileage. The equipment needed for this plant method include an accurate scale and a solvent resistant container of one gallon capacity or greater. The accuracy of these meaurements greatly improves when using a larger volume container-such as two or five gallon capacity. The IF1 tests used a spring scale, purchased as a fish scale. If a spring scale is used, the container should have a handle so that it can be hung on the spring scale. With the container on a level surface, add one gallon of accurately measured water and mark the water level with a sharp line all the way around the container. Measuring Still Residue In the case of still residue, the procedure is very easy. After filling the container with warm still residue up to the one gallon mark, let the residue cool to room temperature. If the volume drops any on cooling, add a little residue to bring it back up to the mark. Weigh the residue. Calculate the net weight of the still residue by subtracting the weight of the empty container from the total weight. Table 10 provides the conversion of percent perc- Le., pounds perc per 100 pounds residue. For example, if net residue weight is 10% pounds, your perc content would be 61 percent. Measuring Cooker Muck Residue Cooker residue is more difficult to accurately measure than still residue because it is a looselypacked, semi-solidhemi-liquid mass. To achieve any reasonable accuracy, tight packing is a must so that there are no air spaces or voids left in the container when filled with muck. This can be done uniformly by the following method: Let the cooker muck cool to room temperature in any temporary (but covered) container. When the muck is cooled, fill the one gallon container about one-quarter full and use the end of a short 2x4 board or other suitable object to pack the muck tightly. Add additional cooker muck and repeat this until the packed level of the muck exactly reaches FOCUS ON DRYCLEANING 3/88

13 the one gallon mark. Weigh the sample, and subtract out the weight of the empty container to get the net muck weight. Find the net weight of your cooker muck in the left column of Table 11; the right column will give you the percent perc left in the cooked residue. Thus, if your net weight is 6.75 pounds, the residue contains 24 pounds of perc per 100 pounds of residue, Accuracy of This Method These conversions were compiled from data derived from numerous samples collected. Some variation may occur, however, when the pure residue (detergent, fats, oils, etc.) varies appreciably from 7.50 pounds per gallon. The accuracy of the cooker residue data is influenced by the amount of water present in the cooked muck and also by the maximum packing density achieved in the container. While this is not a perfect method of determining the perc contained in cooker muck, it presents the only feasible way to make this determination under plant conditions. Any excessive moisture in the cooker muck residue or imperfect packing will lead to falsely high reading for perc content. This method is not applicable to analysis of perc content in residues to satisfy the hazardous waste regulations. The limits of allowable perc content are so low that sophisticated analytical instruments, such as gas chromatograph must be used for such analysis. Reducing Perc Content in Residue IF1 has developed a method that can substantially reduce perc content in residues. While the use of steam-or air-sweep reduces perc content in residues, the water addition into the still or cooker works even better. Water Addition Method The addition of water into the still or cooker follows the final boil down or cook down. After the still or cooker has cooled for at least 15 minutes, the steam is turned on and regular distillation is continued. In stills with an automatic shut-off valve, this process will raise the temperature high enough so that the valve closes, causing interruptions and delays. Eliminate this problem by adjusting the sensor to a higher temperature, or removing the probe and temporarily putting a stopper in the opening. This process also requires careful monitoring of the distilled solvent coming from the water separator. Too much flow will force the solvent out of the water line in U- tube type separators. Although the quality of solvent recovered during the IF1 testing was good, we advise you to collect the solvent in a covered container so that you can periodically check the odor of the solvent. In Still Still residues are liquid. During IF1 tests conducted on stills, after one water addition, the perc content on a weight-to-weight basis was 3.37 percent. This is a 94.1 percent decline from the average 54.7 percent weight-to-weight perc content without adding water. After the second addition of water, the perc content was lowered to an average of 0.36 percent. Following is a step by step procedure that should be used in-plant: Distill and boil down as usual. Let the still cool at least 15 minutes. Add water in an amount equal to the amount of still residue (by volume). Turn on steam. Continue boil down at 25 to 35 pounds per square inch pressure. Continue until water from water separator and from solvent line stops flowing. Cool still 10 to 15 minutes. Add water in an amount equal to one-half the amount of still residue (by volume). Continue to boil down as before. In Muck Cooker Cooker residue is a powder in a liquid. In cookers, the average perc content after cook down was 2.9 percent (weight-to-weight), down 92.9 percent from the average of 41 percent (weight-to-weight) before water addition. The procedure that should be used with cookers follows: Distill and cook down as usual. Let the cooker cool at least 15 minutes. Add 2 quarts of water to cooker. Turn on steam. Continue cook down at 20 to 24 pounds per square inch pressure until water from the water separator and from the solvent line 3/88 FOCUS ON DRYCLEANING 13

14 ~~ ~ ~ INTERNATIONAL FABRICARE INSTITUTE stop flowing. In cookers, one water addition is usually sufficient to reduce perc content to very low levels. Results may vary from plant to plant due to differences in residue constituents, the temperatures of the cooling coils, and the cleanliness of the still's heating coil. Measuring and Reducing Solvent Losses in Cartridges Solvent content in cartridges can be checked in the plant with a simple method: 1 Reweigh the cartridge when additional drying is done with a cabinet, adsorber, or recovery tumbler. Place the cartridge outside the plant in a protected location where it can't get wet and where no one will be exposed to any remaining solvent. Air-dry the cartridge for five to seven days and reweigh it. The original (drained) weight minus the final (outside air-dried) weight will give the solvent loss per cartridge in pounds. For perc, divide by 13.5 to calculate gallons lost per cartridge: for petroleum solvent, divide by 6.5. In case of fluorocarbon solvent (Valclene"), divide by 13.2 In all three operations [perc, petroleum, and fluorocarbon), there are two factors which can influence solvent losses in cartridges, cartridge life and draining time. The draining and/or drying of the cartridge before discarding it increases solvent losses. IF1 recommends draining the cartridges at least overnight and preferably over a weekend. This applies for all three operations, but perc and fluorocarbon plants can go one step further. If a drying cabinet connected to an adsorber is available, the additional solvent recovered-about 0.3 gallons for each cartridge. If an adsorber is available but not a cabinet, a small housing can be constructed that will hold one or more cartridges. The housing can be connected to an inlet duct on the adsorber so that fresh air is drawn over and through the cartridges and then sent to the adsorber. If no adsorber is present, the recovery tumbler can be used to recover solvent from several cartridges at one time if the heating and condensing systems and the fan can be run without rotating the cylinder. This method is practical only if three or more cartridges are run simultaneously: otherwise, losses during tumbler aeration account for a high proportion of the solvent recovered from the cartridges. There are perchloroethylene recovery units commercially available for perc reduction in cartridges. An IF1 Fellowship bulletin evaluates one such unit. Operating this unit for 8 hours, almost all perc was reclaimed from the spent cartridges. Figure 3. Leaky Dampers A leaky damper can have the following causes: Worn dampers. The neoprene gasket material on the face of the damper can become worn, chipped, or cracked from repeated opening and closing. This can cause small unnoticeable leaks. Elongated springs. Some dampers are pushed tight against the vent opefiing by one or more springs. After a period of time, these springs lose their resiliency and cause the damper to leak. Lint build-up around the vent. If the lint gets by the lint bag or lint filter, it can accumulate on the vent opening and on the face of the damper. This lint can prevent the damper from sealing tightly against the vent opening. Mechanical misalignment of the linkage. This problem has the potential to cause the biggest leaks. If a nut or bolt comes loose, as is frequently the case with moving parts, the damper assembly can be forced off center. This would leave a large portion of the vent opening uncovered. Because the problem of leaky dampers can be so costly, it is necessary to periodically inspect them to ensure their proper operation. At least once a week, the ducting leading from the exhaust damper should be removed so that a visual inspection can be made. Make sure that there is no lint buildup and that the damper opens and closes properly. Even if the damper appears to be operating correctly, there can still be small leaks from weak springs or a worn damper. The best way to ensure that your reclaimer is airtight is to place a plastic bag over the mouth of the vent. If the bag inflates when the damper is closed, - 14 FOCUS ON DRYCLEANING 3/88

15 you have a leak that needs to be repaired. See Figure 3. Repairing a leaky damper is simply a matter of replacing worn parts and tightening all the nuts and bolts in the mechanical linkage. Also, any lint that builds up on the damper assembly should be removed. Don t be misled into thinking that a vapor adsorber can reclaim solvent lost through a leaky damper. Unless your adsorption schedule is adjusted so that you are stripping more often, the extra perc will quickly saturate the carbon bed and then pass out of the adsorbers exhaust stack. Leaky dampers can be very costly in terms of solvent losses, and they can add to perc vapor level: in your plant. Misuses of a Vapor Adsorber INTERNATIONAL FABRICARE INSTITUTE Following are most frequent misuses of vapor adsorber: Strip-down (desorption) frequency. The vapor adsorber should be desorbed before the carbon bed becomes saturated with solvent. When the carbon becomes saturated, any additional perc vapor will simply pass through the carbon bed and be lost through the exhaust stack, thereby decreasing the solvent mileage. With excellent operating conditions, a plant may be able to process up to approximately 1500 pounds of cleaning before stripping, using a standard singlebed adsorber with 300 pounds of carbon (this size adsorberwill hold about 4.5 gallons of perc at saturation). For many-or possibly most-plants, approximately 750 to 1000 pounds of cleaning is the most that can be safely done before stripping. (As a rule of thumb, about 1.5 gallons of perc can be held for every 100 pounds of carbon in the adsorber. With an average 40 percent extraction retention and an excellent 90 percent tumbler efficiency, 1500 of cleaning will send about 4.5 gallons of solvent to the adsorber. If a plant has a more typical 80 percent tumbler efficiency, then only 750 pounds of cleaning will send the same 4.5 gallons to the adsorber.) In addition to differences in tumbler efficiency, leaking tumbler exhaust dampers and other factors will dramatically affect stripping frequency. The key to determining stripping frequency for your operation is as follows: begin shortening the intervals between strip-downs until the solvent return is a little less than the manufacturer s rated capacity for your adsorber. Failure to dry out the carbon bed. Failure to dry the carbon bed leads to decreased adsorption capacity and corrosion. Wet carbon will not adsorb perc vapors efficiently or hold as much total solvent. When the carbon is wet, the average adsorber may hold only three gallons of solvent. The remaining perc vapors pass to the atmosphere. Metal corrosion can result from the interaction of wet carbon and perc vapors: carbon acts as a catalyst Figure 4. and converts the moisture and perc to hydrochloric acid. To prevent corrosion, the carbon bed should be dried for 15 minutes after each desorbing procedure. Be sure that no vapors are sent to the adsorber from the cleaning system during this time. Keep both top and bottom dampers fully open to ensure full air flow. Airflow manometer check. Manometers and airflow meters are used to detect obstructions in the carbon vapor adsorber by indicating the pressure at which the adsorber is operating. The manometer, found on older models, has a plastic graduatedmeter. The airflow meter, found on newer models, has a dial that is easier to read. Failure to check the manometer or airflow meter daily can result in improper air flow into the carbon vapor adsorber. Consult your owner s manual to determine the normal manometer or airflow level. A lower-than-normal airflow reading means that the adsorber s exhaust duct or damper is not functioning as it should. Clean the ducts, readjust the dampers, or replace the damper gaskets as necessary. Once these have been checked and cleaned, normal airflow should be restored. Exposure to petroleum solvent vapors. If petroleum solvent vapors are adsorbed by your carbon, the bed s capacity to adsorb perc is reduced. This is commonly called poisoning the carbon bed, Petroleum solvent will not strip or desorb at the steam temperature used for perc, but requires much higher temperatures. The vapors remain in the carbon bed and decrease the amount of carbon available to adsorb perc vapors. Be careful not to use detergents that contain petro- 3/88 FOCUS ON DRYCLEANING 15

16 leum co-solvents as a dilutent. Only drycleaning detergents designed for perc systems should be used since they should contain no petroleum co-solvents. Overheating the carbon bed. Overheating the carbon bed can be a problem when several pieces of equipment are exhausted into the adsorber, with drying cabinets usually being the main culprit for this problem. Whether drying garments or cartridges, it usually takes several hours of heating in the cabinet to evaporate the perc from heavy garments or spent cartridges. It is best not to use the drying cabinet for many hours at a time without turning it off for short periods to allow the carbon to cool down. Without sufficient cooldown, the carbon becomes hot and loses its capacity to adsorb perc vapors. Perc vapor loss starts once the bed reaches approximately 100" F, no adsorption takes place and all perc vapors are exhausted to the atmosphere. out the water drain line. In Figure 4 we show some corrosion inside of the separator; this is relatively small, particularly for a separator that has been in use for 13 years. To check for unwanted material in your separator, drain the solvent and water by removing the drain plug that is on the bottom of most separators. If your separator has a removable top, you can look directly inside for any buildup on the base. Flush the separator with water. When flushing with water, be sure to lock the drain line going back to your base tank first. You can do this from the inside by using a cork or other stopper. Using a bent coat hanger or something similar, fish into the solvent and water drain lines from the inside of the separator to see if there are any lint accumulations. Next, fish down to the vent lines to see if there is any blockage. If you have a water separator which is a sealed unit, such as the one shown in Figure 5, you'll have to do all of your work from the outside using a coat hanger. Flush the separator again with water to remove any sediment lying on the bottom-be sure to block off your solvent return line to the base tank. If you run your water drain line from the separator directly to a sewer line, you can break this line close to the separator itself and install reducing coupling. This is also shown in Figure 5. By doing this, you can easily collect the water coming out of your separator to check for perc, or you can catch some of the water on your hands to smell for the presence of perc in the water. After checking and cleaning out your water separator, put enough solvent back into the separator to come up to the level of the solvent outlet line back to the base tank. Then pour a small amount of water into the separator-be sure to pour this into the middle of the separator, and not between the solvent outlet line and the baffle closest to it. Water Separators In perchloroethylene plants, water separators are used on recovery tumblers, cookers or stills, and vapor adsorbers. Water separators are simple devices with three pipe connections, one or two inside baffles, and no moving parts. So what can go wrong? Nothing usually-unless you have solid material inside the separator that doesn't belong there. These particles consist of large flakes of rust or lint. Lint, of course, would be most likely to occur in a tumbler water separator, while rust is more likely to happen on still/cooker or adsorber water separators. Either of these materials can block solvent drain line or the vent line(s) which equalize pressure inside of the separator. When this happens, the solvent level may rise high enough inside of the separator for it to flow Figure FOCUS ON DRYCLEANING 3/88

17 Dryers and Recovery Tumblers Among other factors contributing to the efficiency of recovery is cleanliness of the lint bag. Excessive lint build up can reduce airflow and result in tears which allow lint to get through and deposit on condenser and even heating coils. In such cases, the efficiency of the recovery-dryer is seriously reduced because of the insulating effect of the lint on the heat transfer surfaces. The proper airflow rate is also important and it can be impaired by a lint accumulation in various parts of the recovery unit, or by the faulty installation of the fan motor, such as a fan motor rotating backwards. In both cases, the recovery unit will function, but very inefficiently. This condition has been found on both new tumblers and those that have been running several years. To prevent this problem, check to see that the fan motor is rotating in the proper direction (there should be an arrow on the fan motor casing that indicates proper rotation]. This condition not only increases the required time for drying but can also force saturated air out of the air inlet damper during the deodorizing cycle. Overloading. Overloading a tumbler reduces overall garment movement during drying and restricts airflow. If you overload by 25 percent to 30 percent, recovery time must be increased by approximately 6-8 minutes or efficiency will drop. The most important temperature during drying is that of the air after it leaves the cylinder but before it reaches the condenser. This is stack temperature and it should be in the range of 135O-145O F. If you don't have a thermometer built into your tumbler which measures this temperature, you should install your own-either a common bulb type or a metal dial type, Figure 6 shows a bulb thermometer installed in the tumbler air passageway after the cylinder and before the condenser. Recovery in petroleum recovery tumblers reacts to the same factors as involved in perc tumblers, However, since petroleum is combustible, particular care must be taken to follow manufacturer recommendations explicitly. Be sure you are using the proper condensing water temperature, heating steam pressure, and air flow rate before you begin. multiple relaxation and depression of the gasket (as happens at the cleaning machine door] or where there is very hot solvent in contact with the gasket (as occurs at a cooker door]. See Figure 7. While. adjustment at the hinge of a door will usually offer enough additional compression to stop a leak which has just begun, the gasket for that door should be replaced as soon as possible. If you're replacing a cooker door gasket, be sure to get the recommended one-these gaskets are specially designed to withstand extreme conditions. Ductwork in poor condition can be significant source of vapor leaks into the cleaning room area. Ducting from the cleaning machine, recovery tumbler, or adsorber pickups should be solid and noncorroded. In addition, all ducts should be firmly supported from the wall or ceiling by brackets or rod hangers. You should not depend on the rigidity of the duct itself to hold the entire system in place. Vibration from the cleaning machine and tumbler will cause the duct to shift and change alignment, allowing the seams to open. Duct joints should be securely fastened together, either with pop rivets or sheet metal screws. This gives mechanical rigidity to the joint, but it is not the only step to take. A good quality duct tape should be wrapped around the joint at least twice to seal any leakages which could result from a poor fit at the joint. Deteriorating duct joints are once of the most common problems that we have seen in plants visited. Miscellaneous Gaskets and Pipe Joints. If you find leaks at pipe fittings or unions, tighten them and then check after serveral more runs. If you still have leaks, you may need to use either pipe cement or tape or replace the fitting. Liquid solvent leaks are generally caused by loose pipe joints and/or defective door gaskets. Most loose pipe joints (and unions] can be corrected by either tightening the pipe or by the application of a thread sealer, such as pipe dope or Teflon tape. Door gaskets become hard and overly compressed with age and use. This is particularly a problem where there is Figure 7. Example of leaking gasket on washer door. 3/88 FOCUS ON DRYCLEANING 17