CONVERSIONS TO NON-CFC REFRIGERANTS SIMPLE, INEXPENSIVE

May 29, 1995
Richard A. Corbett R.A. Corbett Engineering Houston The U.S. Environmental Protection Agency (EPA) regulations that ban the production of chlorofluoro-carbon (CFC) refrigerants after Dec. 31, 1995, are prompting refinery and petrochemical plant operators to convert existing refrigeration systems to operation on non-CFC refrigerants. Many of the systems that use CFC refrigerants can be converted without significant changes to the major components of a refrigeration system, and at a cost
Richard A. Corbett
R.A. Corbett Engineering
Houston

The U.S. Environmental Protection Agency (EPA) regulations that ban the production of chlorofluoro-carbon (CFC) refrigerants after Dec. 31, 1995, are prompting refinery and petrochemical plant operators to convert existing refrigeration systems to operation on non-CFC refrigerants.

Many of the systems that use CFC refrigerants can be converted without significant changes to the major components of a refrigeration system, and at a cost substantially lower than replacing the system.

Converting an existing system will relieve concerns about the availability and cost of CFC refrigerants when production of them ceases at the end of this year. And conversion to a non-CFC refrigerant can relieve concerns about the new rules for handling refrigerants and servicing equipment.

A recently completed conversion study of three typical process refrigeration systems installed in a major Petrochemical complex on the U.S. Gulf Coast, typifies the analyses and procedures required to successfully convert a system to non-CFC operation. The study also provides insight into some potential refrigerant alternatives that are effective replacements for CFCs.

Also highlighted are recommendations and limitations for conversion of some common types of refrigeration systems used in refining and petrochemical processing.

THE REGULATIONS

In 1987, 24 countries and the European Union met in Montreal to establish a schedule for phasing out production of ozone-depleting chemicals. This "Montreal Protocol" led EPA to establish a phaseout schedule for ozone-damaging materials as part of its implementation responsibilities under the Clean Air Act Amendments of 1990 (CAAA).

The CFC refrigerant phaseout rules are covered under CAAA Section VI, Stratospheric Ozone Protection. EPA's phaseout, which began in 1992, will be complete on Dec. 31, 1995, when production of CFC refrigerants will cease completely.

EPA rules also govern the handling of halocarbon refrigerants, the servicing of refrigeration systems, and the qualifications of servicing personnel. Section 608 of the CAAA, effective July 1, 1992, prohibits individuals from knowingly or intentionally venting into the atmosphere ozone-depleting compounds used as refrigerants.

To prevent venting, EPA also requires certified refrigerant recovery equipment to be used for system servicing. To ensure proper handling of refrigerants and servicing of refrigeration systems, EPA, after Nov. 14 1994, ruled, per 40 CFR, Part 82, Subpart F, that all refrigeration technicians must become certified by passing an EPA-approved test. Under the same ruling, the sale of refrigerants is restricted to certified technicians.

OPERATOR CONCERN

Because process-plant refrigeration systems can use large amounts of CFC refrigerants-some systems require more than 20,000 lb of refrigerant charge-plant operators fear there will not be adequate supplies after the production ban. And many of these refrigeration systems are critical to process operations.

Although it is expected that, after CFC production ceases, reasonable inventories will remain as a result of recovery and reclamation activities, those inventories will naturally dwindle. Plant operators will have to compete with other refrigerant users, such as automobiles and home appliances (which use R-12, in particular) that usually cannot be converted to non-CFC refrigerants.

The refrigerants that will not be produced are R-11, R-12, R-113, R-114, R-115, R-502, and some others (Table 1)(17737 bytes). It should be noted that, although the use of CFC refrigerants will not be prohibited, none of these refrigerants will be manufactured any longer.

With the end of CFC production looming, the costs of CFC refrigerants have already risen substantially. R-12, a common CFC refrigerant used in many process systems, presently costs about $6-7/lb in bulk, compared to less than $2/lb only 5 years ago, and the cost is likely to rise further after production ceases.

Part of the cost increase is associated with special excise taxes levied on CFCs (currently $5.35/lb on R-12) to encourage use of non-CFC compounds. But market forces, driven by expected supply constraints, are also pushing up prices.

Demand for CFC compounds also will decline after production ceases. In 1992, refrigeration and air conditioning accounted for about 40% of total demand for CFCs (Fig. 1).2(43437 bytes)

AFFECTED REFRIGERANTS

EPA has designated CFC refrigerants Class I substances which, by definition, are known to significantly cause or contribute to ozone damage. To compare the relative destructive effect each compound has on stratospheric ozone, each CFC refrigerant was assigned an ozone depletion potential (ODP) factor (Table 1)(17737 bytes). Class I refrigerants have ODP levels 0.2 or higher.

The names and chemical formulas for some of the common CFC refrigerants (Table 1)(17737 bytes) used in process plants are:

  • R-11, CC13F, Trichlorofluoromethane

  • R-12, CC12F2, Dichlorodifluoromethane

  • R-113, CC12F-CC1F2, Trichlorotrifluoroethane

  • R-114, C2C12F4, Dichlorotetrafluoroethane

  • R-115, CF3CF2C1, Chloropentafluoroethane

  • R-502, Azeotrope of R-115 and R-22.

R-22, CHC1F2 (chlorodifluoromethane), is designated a Class 11 substance and a hydrochlorofluorocarbon (HCFC), not subject to the production ban at the end of this year. R-22 and other Class 11 HCFCs are subject to gradual phaseout, with production ceasing in 2030 at the present tentative phaseout schedule.

HCFC production phaseout occurs later because hydrogen in the compound makes HCFCs much less stable than CFCs. Their shorter lives in the atmosphere reduce their potential for causing damage. The ODP for R-22, for instance, is 0.05.

REPLACEMENT STUDIES

A thorough investigation of system performance and potential refrigerant alternatives provided important information and data that led to the decision to convert three systems to a non-CFC refrigerant. Propane was the eventual choice to replace the R-12 in the three systems, but other non-CFC refrigerants, such as ammonia and some new halocarbons that do not contain chlorine, also can be good choices (Table 2)(43515 bytes).

Three butadiene chilling systems, installed and operating at a major U.S. Gulf Coast petrochemical complex, were analyzed to decide whether they should be converted to non-CFC operation or replaced with new equipment. The systems are used to chill liquid butadiene to maintain low vapor pressure for storage.

Each of the systems uses a screw compressor, a shell-and-tube-type butadiene chiller, and an air-cooled refrigerant condenser (fin-fan type),and each includes required system vessels (receiver, oil separator, etc.). The systems are typical of refrigeration systems used in many process applications.

Two of the systems, connected in parallel, provide a total of 1,430,000 BTU/hr (119 tons) of refrigeration capacity when chilling 552,000 lb/hr of butadiene from 40 to 35 F. The third system is similar in design, but is rated at 780,000 BTU/hr (65 tons) of refrigeration capacity and (,hills 250,500 lb/hr of butadiene from 42 to 36 F.

A simplified flow diagram, representative of all three systems, is shown in Fig. 2.(60410) All three systems contain the same components, except for the compressor lubricating oil cooling system. Details of the oil cooling methods are important because the oil cooling systems had to be modified for the conversion to a non-CFC refrigerant.

For two of the systems, lubricant is cooled by thermal siphon oil coolers. The oil coolers are shell-and-tube-type heat exchangers with liquid refrigerant from the refrigerant condenser on the tube side and lubricating oil on the shell side. Heat from the oil evaporates the refrigerant, and the resulting refrigerant vapors are returned to the condenser.

Oil cooling for the third system is accomplished by directly injecting liquid refrigerant into the compressor casing. Injection is controlled by temperature, and is fed into the compressor at a rate such that all refrigerant is evaporated before leaving the compressor.

CANDIDATE REPLACEMENTS

Because all of the systems use positive displacement compressors, and because they all operate at moderate temperature (35 F.), four candidate refrigerants were selected:

  • R-134A, a hydrofluorocarbon compound with similar properties R-12 and not subject to the production ban

  • Ammonia (R-717), a refrigerant that has been used for many years in industrial systems with positive displacement compressors

  • Propane (R-290) and propylene (R-1270), two hydrocarbons commonly used for process refrigeration applications.

R-22, the HCFC, was considered initially, but because it is subject to gradual phase out and an eventual production ban after 2000, this candidate was eliminated. R-22, however, may be a suitable candidate for conversion of other systems.

Table 2 (43515 bytes) shows a comparison of the chosen candidates, and other potential replacement candidates. Important properties are shown for each.

ANALYSES

Each of the systems was evaluated for performance on each of the candidate refrigerants (R-134a, R-290, R-1270, and R-717). The steps used in the evaluation included:

  • Evaluate all components for design pressures.

  • Rate compressors on each candidate refrigerant.

  • Rate chiller and condenser on each refrigerant.

  • Conduct a heat and flow balance for each refrigerant.

  • Evaluate vessels for proper sizing.

  • Check pipe sizes for flow and pressure drop.

  • Examine materials for compatibility with new refrigerants.

  • Define new operating set points with each refrigerant.

Checking the pressure ratings of all system components shows which refrigerants can be suitable replacements based on the condensing pressure of the replacement candidates. Table 2 (43515 bytes) shows condensing pressures for each of the refrigerants listed.

All components for the systems studied had maximum design operating pressures of at least 300 psig. Therefore, propane, propylene, and ammonia could be used, even though their condensing pressures are significantly higher than that of R-12.

The system compressor, or compressors, must then be rated for operation on each of the selected candidates. The rating is carried out at the original design operating temperatures.

The original temperatures are used so that the system will provide the same product or reaction temperatures required by the process. The condensing temperature is the same because condensing temperature is based on the ambient conditions of the local area.

For the systems studied, compressor capacity remained essentially the same for operation on R-134a. Compressor capacity, however, increased substantially with propane, propylene, and ammonia. Table 3 (30872 bytes) shows the compressor capacity comparison.

When the butadiene chillers and refrigerant condensers were rated for each refrigerant, it was found that all of the exchangers would be adequate for any of the refrigerants. Table 3 (30872 bytes) shows the heat exchanger capacity comparison.

When the ratings were conducted for the two compressors with oil coolers, it was found that the coolers were adequate for operation on R-134a, but inadequate for operation on propane, propylene, or ammonia. It was also determined. that, on the refrigerant injection-cooled compressor, direct injection of propane or propylene for oil cooling would not be advisable because of the solubility of propane or propylene in the lubrication oil.

Although the compressors produced higher capacity for propane, propylene, and ammonia, much of that capacity gain was not available because of capacity limitations of the butadiene chillers.

It can also be seen in Table 3 (30872 bytes) that, although the compressors could produce more capacity, they also had higher power requirements. The capacity gain, therefore, was also limited by the installed drive motor size.

From a performance standpoint alone, ammonia and R-134a looked like logical choices, based on the power required per ton of refrigeration produced. But ammonia becomes less attractive because the compressor would have to run at substantial partial-capacity conditions.

A complete heat and flow balance was conducted so that control valves, service valves, and piping could be analyzed for pressure drop and flow capacity. All control valves, hand valves, and piping were shown to be adequate for any of the candidate refrigerants.

The refrigerant receivers were checked to determine whether they could contain the charge requirements for each refrigerant. Receiver sizing was shown to be adequate for all of the candidates.

LUBRICATING OIL

In most refrigeration systems, the compressor lubricating oil circulates through the system with the refrigerant. Therefore, the different refrigerants require different lubricating oil properties so that the oils properly lubricate the compressor, and at the same time, circulate well through the system for recovery and return to the compressor.

With screw compressors, a fairly large amount of lubrication oil is injected into the gas stream, thus requiring oil separators to minimize oil carryover into the refrigeration system. Oil separator efficiency is a function of both lubricant and refrigerant density.

Therefore, it is extremely important to analyze oil separation efficiency for each oil/refrigerant combination. The oil separators were evaluated for each candidate refrigerant, and existing oil separation was found to be adequate for all candidates.

Oil carryover is also an important consideration for systems that use reciprocating compressors, particularly those that use compressors designed specifically for refrigeration service. Oil carryover is much less important for systems that use centrifugal compressors or process-type, non-lubricated reciprocating compressors. Centrifugal and non-lubricated reciprocating machines normally have very low oil carryover levels.

MATERIALS

A detailed evaluation of all materials of construction of the system is important to determine the compatibility of all materials with the refrigerants and with the required lubricating oil.

For instance, copper materials are adequate for use with halocarbon refrigerants, but they cannot be used with ammonia because ammonia will attack copper if there is even a slight amount of water present. Although propane and propylene will not harm copper, plant safety considerations usually prohibit copper in pressure-containing parts.

Elastomers used for seals and gaskets in refrigeration systems may not be compatible with the refrigerants and lubrication oils. Each system component must be examined for incompatible elastomers so that seal swelling and gasket deterioration will not occur.

Viton, for instance, will undergo significant change with R-134a. however, Buna N shows no degradation. Fortunately, most of the elastomers commonly used in process plants will provide acceptable service with hydrocarbon refrigerants and ammonia.

CONTROLS

When all other steps of the evaluation are completed, all control components and operating controls must be adjusted to set points required for the new refrigerant. In most cases, this will simply require readjustment.

Control adjustments are required because operating pressures change for each candidate refrigerant. And because refrigerant flow rate depends on the refrigerant selected, flow control components must be readjusted.

An important component to evaluate is the level controller that controls liquid refrigerant level in the refrigerant evaporator (usually a heat exchanger). Depending on the refrigerant, the existing floats or displacers used in the level controls must be checked for proper operation with a refrigerant of substantially different density.

In some systems, control components may have to be replaced because of inadequate adjustment range, and because the mass flow rate of the new refrigerant is significantly different from that of the existing refrigerant. Level control valves and thermostatic expansion valves, in particular, may require different control elements because of widely different refrigerant mass flow rates.

For the studied systems, the refrigerant mass flow rate was about 16,000 lb/hr for R-12 and about 14,000 lb/hr for R-134a. But the mass flow rate was only about 7,000 lb/hr for propane and propylene, and only about 1,700 lb/hr for ammonia.

Control adjustment was all that was necessary for the systems studied because the control components and instruments all had adequate adjustment range. However, this may not be the case for other systems.

COST COMPARISON

The information and data gathered from the study of the three systems showed that all three units could be converted to any of the candidate refrigerants with only minor component changes. When conversion costs and performance for all candidate refrigerants were compared, propane was chosen.

Conversion costs varied depending on the refrigerant selected. Regardless of the refrigerant chosen, none could be simply "dropped" into the system. This is generally the case for all refrigeration systems considered for conversion.

As stated previously, lubricating oil circulates through a system with the refrigerant, with different lubricating oil properties required for different refrigerants. Therefore, one the most important requirements of any conversion is to thoroughly remove the existing lubricating oil.

In the case of R-12 and other halocarbon refrigerants, mineral-oil-type lubricants are used. However, R-134a requires the use of a polyol-ester synthetic lubricant, and the lubricant will not work properly with even traces of mineral oil in the system.

Lubricants required for propane and propylene are polyalkyline-glycol types, and they, too, do not work well when mixed with small amounts of mineral oil. Mineral oils usually are acceptable for use with ammonia.

Conversion to R-134a would require the least amount of system change because its properties are very similar to R-1 2. Therefore, most of the conversion work would involve flushing out the old oil, readjusting controls, and resetting instrument set points.

As stated previously, conversion to propane, propylene, or ammonia would require new oil coolers, in addition to flushing oil and readjusting controls.

Conversion to propane or propylene also would require the addition of emergency shut down (ESD) valves to isolate the compressor in case of fire. The compressor casings are constructed of cast iron, which usually is not acceptable for hydrocarbon service unless an ESD system is installed.

The concern is that cast iron will crack during a fire, allowing a large amount of hydrocarbon refrigerant to escape. The ESD valves prevent this. The control systems on each unit also would have to be modified to operate the ESD va e .

PROPANE CHOSEN

A cost comparison shows that, even though R-134a required the least system changes, the refrigerant cost more than offset the modification costs for the other refrigerants (Table 4)(22039 bytes).

R-134a was estimated to cost about $6.00/lb at the time of the study. The estimated charge of 11,000 lb each in two units, and 7,000 lb charge in the third unit, resulted in a total refrigerant cost of about $174,000.

The total amount of either propane or propylene in the three systems was about 13,500 lb at about $1.50/ lb on site. The lower cost of refrigerant (approximately $20,000) more than offset the higher base conversion cost of about $160,000. The higher equipment costs were associated with new oil coolers and the ESD systems required with propane or propylene.

Basic conversion costs for ammonia were slightly less than for propane or propylene because the ESD system would not be required. Ammonia refrigerant costs were substantially lower than propane or propylene at about $0.30/lb, for a total refrigerant cost of about $2,500.

Clearly, ammonia would have been the choice from an economic standpoint alone. However, plant engineering personnel were concerned that procedures for the safe handling of ammonia would have to be developed because the plant had no such procedures for handling, storing, or disposing ammonia.

Because safety procedures at the plant were well established for propane and propylene, the final choice was limited to these two candidates. Propane was selected because it was readily available at the plant and because it had much better efficiency, in terms of bhp/ton, than did propylene (Table 3)(30872 bytes).

The rules for refrigerant handling, recovery, and purchasing halocarbon refrigerants do not apply to propane. Therefore, no additional certification of operators or maintenance personnel is required. Of course, propane must be handled in accordance with EPA rules established for hydrocarbon emissions, and with handling procedures already well developed at the plant.

The overall cost to convert all three units to operation on propane, at $160,000, is significantly less than the cost of replacing the units. The estimated replacement cost for the three systems, designed for operation on any non-CFC refrigerant, could easily exceed $750,000 at today's equipment costs.

Conversion, rather than replacement, was obviously warranted. All three conversions are currently in progress. Completion is expected by the end of the year.

REFERENCES

  • U.S. Regulatory Update, Du Pont Chemicals, Wilmington, Del., November 1992.

  • "CFC Alternatives Making Progress," Chemical Marketing Reporter, Jan. 2, 1995, p. 3.

  • ASHRAE Handbook, Fundamentals, American Society of Heating, Refrigerating, and Air-Conditioning Engineers Inc., Atlanta, 1993, Chapter 16.
    Copyright 1995 Oil & Gas Journal. All Rights Reserved.