R8515 ESE GEM ABSTRACT ENGINEERING ANALYSIS OF THE CONTROLLED ATMOSPHERE STORAGE by David Gurevitz An engineering analysis of some of the unit operations in sev- eral phases of controlled atmosphere (CA) storage design, and also some new approaches and suggestions are presented. This thesis deals with the CA storage of apples, but most of the observations are applicable to other types of produce. The methods for control of the gas atmosphere have been critically reviewed and design information to make possible the selection and operation of equipment in a CA storage are presented. A new way to remove carbon dioxide using a selective membrane is suggested. A cost analysis of the various methods to remove excess carbon dioxide and oxygen in order to control the atmosphere is presented. A method showing which system to use to control the atmosphere under particular conditions is outlined. As there are many alterna- tives to reduce carbon dioxide and oxygen, many different combinations can be used to control the atmosphere. As a guide to the choice of the most suitable method of controlling the atmosphere some preliminary elimination of these alternatives can be made easily according to the .fl.;~ .~.u m ‘1 an 0 A; huh. David Gurevitz degree of tightness of the room and the management practices (single or multiple opening of room during season) used. After this elimination step, the most economical combination among the re- maining can be chosen from tables of the cost data for CA operation (Tables ZJ- 3, 4, 5 and 6). A critical literature review of the basic problems involved in using vapor barrier, gas seal materials and their location within the CA enclosure was made. The problems of moisture migration through the insulation, the effect of weather, reversible vapor flow and room operational conditions were discussed. A discussion of other possible solutions to water vapor transfer, CA gas tightness and pressures and temperature cycles in the CA storage is presented. A new approach to precooling and air distribution in a CA apple storage room is suggested. In this approach all the spaces between the pallet boxes will be closed and the stacking will be arranged in such a way that the air will be forced to flow through the void spaces between the apples in the pallet boxes. Suction fans distributed uniformly on a false ceiling will be used to move the air through the system and give even distribution in the stack. The most important advantage of this method to CA operation is that there are no spaces between the boxes in the room and more apples can be stored in the same size room. There is a possible saving on storage construction David Gurevitz costs per bu. The time to reduce the oxygen level in the room is shorter than in a regular CA. room and the gas consumption using burner units to reduce the oxygen is lower. This study was based on a theoretical analysis and since no experimental work was done at this stage it presents only a speculation of a new air distribution method. ENGINEERING ANALYSIS OF THE CONTROLLED ATMOSPHERE STORAGE By David Gurevitz A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOC TOR. OF PHILOSOPHY Department of Food Science 1966 (‘7) Copyright by DAVID GUREVITZ lap? Dedicated to my parents 35 merr. comm,- inIEI‘ES ACKNOWLEDGEMENTS The author wishes to express particularly his appreciation to his major professor, Dr. I. J. Pflug for his inspiring advice, help and constant encouragement throughout this study. Grateful appreciation is extended to Prof. L. J. Brazler, Dr. T. I. Hedrick, Dr. J. S. Boyd and Dr. R. F. Gonzalez who served as members of the committee, reviewed the manuscript and made constructive comments. Thanks are expressed to the Whirlpool Corporation, for their interest and support of this project including the graduate assistant- ship that made this study possible. To Dr. B. S. Schweigert, Chairman of the Food Science Department, I should like to express my appreciation for making available the graduate assistantship. Grateful acknowledgement is extended to Mr. C. Borrero and Mr. K. Fox for their patience and help in editing the manuscript. iv LIST OF LIST OF LIST OF .\'O.\lE.\'( SECIlOi SEC T10? 3.: TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES LIST OF APPENDICES NOMENCLATURE SECTION 1. INTRODUCTION SECTION 2. THE CONTROL OF THE ATMOSPHERE . 2A. 2B. 2C. 2D. 2E. 2F. 2G. 2H. 21. 2J. SECTION 3A. Introduction . Restricted Ventilation . Removal of Excess Carbon Dioxide 2C1. Introduction. . . . . 2C2. Physical absorption processes 2C3. Chemical absorption processes . 2C4. Removal by use of molecular sieves . 2C5. Removal by using a selective membrane . Oxygen Reduction . ZDl. Fruit or product generated atmosphere ZDZ. The addition of nitrogen or carbon dioxide 2D3. Hydrocarbon fuel burner . 2D4. External gas generator External Gas Generator Complete Systems for the Control of the Atmosphere CO; and 02 Measurement and Control Systems Humidity Control Air Purification Engineering Cost Analysis Study of CA Storage Systems . . . . . . . 3. GENERAL CONSIDERATIONS IN THE DESIGN AND SELECTION OF VAPOR BARRIER AND GAS SEAL MATERIALS Introduction . 3A1 . General Page vii viii xi 107 112 127 127 127 SECI SECI SEQ- APp Lin 3B. 3C. 3D. 3E. SECTION 4A. 4B. 4C. SECTION 5A. 5B. 5C. SD. SE. 3A2. Water vapor barriers 3A3. Thermal insulation - Moisture Migration Through Materials . The Effect of Weather and Reversible Vapor Flow . . . . . . . . Gas Sealing the CA Room . . . Analysis of Design Principles and Possibilities 3E1. One direction vapor flow 3E2. Reverse flow 4. OPERATIONAL FACTORS AFFECTING GAS LEAKAGE IN CA ROOM PERFORMANCE Pressure and Temperature Cycles in the CA Storage Room - - - . . - - Eliminating Pressure Changes and Devices to Prevent Room Damage Due to Pressure Changes- Measuring the Gas Tightness of CA Storage Rooms 5. PRECOOLING AND AIR DISTRIBUTION IN A CA APPLE STORAGE ROOM . Introduction . . General Design of New Method of Storage . . General Outline for Analytically Determining the Value of the Variables in the "Packed- Bed" Type Storage Room Operating System . . The Method to Calculate Design Data for the. "Packed- Bed" Room--Precooling and Air Distribution . . 5D1. The approximate calculation of T1 and h 5DZ. Calculation of the "exact" cooling curve 5D3. The determination of the air velocity v needed and chm . . 5D4. The determination of maximum APB needed for air distribution . 5D5. Evaluation of the total refrigeration load Advantages, Disadvantages and Discussion of the New Approach SECTION 6. DISCUSSION AND CONCLUSIONS APPENDICES LITERATURE CITED vi Page 128 129 130 132 134 138 138 140 147 147 155 158 161 161 163 165 167 168 169 172 173 175 176 178 186 201 Tab' .3 C 7 2C5 7E 2F- )H Table 2C2-1 2C3-l 2D1-1 2F 1 2J1 2J2 2J-3 2J4 2J-5 2J6 LIS T OF TAB LE S Mass transfer rate in packed absorbers-- list of references . Mass transfer rate in chemical reaction packed absorbers--list of references Time required for CA storage room to reduce oxygen level to 0.030, carbon dioxide level to 0.025, as a function of type of carbon dioxide absorption system, leakage rate and room fullness Possible combination for complete system for the control of the atmosphere in a CA room . Annual cost for carbon dioxide removal (cents per bushel) Annual cost for oxygen reduction (cents per busher for 4 pull down periods) . . Annual cost in cents per bushel for combination of carbon dioxide removal and oxygen reduction-- small scale operation . . ..... Annual cost in cents per bushel for combination of carbon dioxide removal and oxygen reduction-- large scale operation . Annual cost in cents per bushel for a several existing complete CA systems Estimated annual cost in cents per bushel for a "polastream" modification and dry lime scrubbing as CA system for the whole season. vii Page 19 61 82 114 115 116 118 120 121 Fig: 3C2 7C2 3C3 2C3 2C3 2C: 3C. 21) Figure 2C2-1 2C2-2 2C3-1 2C3-2 2C3-3 2C3-4 2C3-5 2C4-1 2C4-2 2C5-1 2D2-1 2D3-1 2D3-2 2D3-3 2E 1 2E2 2G 1 LIST OF FIGURES Simple basic design of tower with full jet spray nozzle for water scrubbing ............... Cross section diagram of water absorber for CA fruit storage ...... Barrel type absorption unit . . . Brine spray absorption unit. Schematic diagram of the M. S. U. absorption system ......................... Dry lime scrubber for CA apple storages Diagram of the Sulzer scrubber Schematic diagram of the CA system. . Flow diagram of Atlantic Research Corp. controlled atmosphere generator. ........ General design of C02 diffuser using selective membrane ..... "Polastream" refrigeration system ....... Oxygen vs. carbon dioxide from combustion in air Effect of room leak on oxygen pulldown time using Arcat burner ...................... Oxygen controller for CA storages . .......... Atmospheric control factors affecting the storage. . Tectrol atmospheric generator schematic . Schematic drawing of C02 and Oz instrumentation system for multiple CA rooms ............. viii Page 12 13 26 26 27 32 39 49 51 56 64 69 71 74 77 78 92 l .' . 1:1- , ~18- Figure 2H 1 2H-2 3C1 3C2 3E-1 4A1 4B-1 4B2 5B1 5B2 5B-3 5B4 5B5 Blow through vs. draw through . Schematic sketch of jacketed cold room Wetting and drying potentials for CA storage structural components ......... . . Histogram of diurnal average vapor pressures Features of design for moisture control within CA storage structural components . . . ..... Trap that allows water to flow from absorption section to desorption section without gain or loss of storage room gas. . . Safety device to prevent excessive pressure in the CA room Relief valve to prevent excessive pressure in CA room ........ Schematic drawing of "packed-bed" room-- side view ..... Schematic drawing of "packed-bed" room-- top view . The use of an extra pallet every second row to prevent horizontal air channeling. ........ Short circuit air flow in regular room precooling during loading period ...... . ........ Precooling during loading- - " pac ked-bed" approach ...... ix Page 98 100 133 135 144 154 157 157 164 164 166 166 166 LIST OF APPENDICES Comparison of Annual Cost of Operation of Orsat and Instrumentation System COZ Removal Costs Oxygen Reduction Costs Complete Systems Annual Costs Page 186 190 194 197 CRF AC La Ga La NOMENCLATURE Section 2 total amine concentration, mols/ volume annual cost, $ quantity of fruit in storage, bushels (bu.) controlled CO; level in storage, cu. ft. COz/cu. ft. atmosphere arithmetic average (top and bottom of column) normality of carbonate capital recovery factor lb. mol of solute cu. ft. /soln. log mean concentration difference, diffusivity of solute gas in liquid, sq. cm. /sec. fraction of total base present as carbonate gas rate, lb. /(hr. )(sq. ft.) height of packing Henry's law constant, (1b. mol/cu. ft.)/atm. interest rate as a decimal liquid film coefficient, 1b. mols/(hr. )(cu. ft.) (lb. mol/ cu. ft.) overall gas coefficient, lb. mols/(hr. )(cu. ft. )(atm. ) overall liquid coefficient, 1b. mols/(hr.)(cu. ft.) (lb. mol/cu. ft.) liquid rate, lbs. /(hr. )(sq. ft. tower cross section) xi :3 Na Na {OH Na Na (OH—5 Xg liquid rate, lb. mols/ (hr. )(sq. ft.) prospective salvage value, $ amine concentration of the solution, g. mols/ liter unit life, number of years lb. mol of solute transferred per hr. moles transferred, lb. mols/(hr. )(sq. ft. of tower cross area) sodium normality arithmetic average (top and bottom of columns) normality of hydroxide partial pressure of soluble gas initial cost, $ respiration rate, cu. ft. 02 consumed/ 1000 bu. day solubility of C02 in water under a pressure of one atm. of CO; time temperature, °C - degrees Centigrade °F - degrees Fahrenheit air velocity storage room volume (gross), cu. ft. vapor pressure tower volume, cu. ft. oxygen diffusion ratio, cu. ft. 02 added/cu. ft. CO; removed weight of water mole fraction of gas in liquid phase xii Fri.) '74 Yt leakage or ventilation rate, air changes (based on empty room volume) 24 hr. oxygen level in storage, cu. ft. 02/ cu. ft. atmosphere oxygen level in storage at time t V space factor Z = B , cu. ft. /bu. empirical constant given for different packing materials carbonation ratio, mol COZ/ mol amine liquid viscosity ionic strength liquid density, lb. /cu. ft. Section 3 area of cross section of flow path, sq. ft. dry weight of hygroscopic material, lbs. length of flow path, in. initial equilibrium moisture content of material, lb. /lb. allowable equilibrium moisture content of materials, lb. / lb. vapor pressure difference, in. of Hg time of transition, hrs. total weight of vapor transmitted, grains moisture storage capacity, grains permeability, grain in. /(sq. ft. )(hr. )(in. Hg vapor pressure difference) xiii Kr" Bi 1p :11 Q '11 {3" Bi AP Section 5 cross .section of bed (stack) the radius of the apple (sphere) apple (sphere) surface area per unit bed area specific heat: C specific heat (at constant pressure) of apple PS CPA specific heat (at constant pressure) of air defined in Eq. 5D2-6 time required for the asymptote of a cooling curve to cross one log-cycle, the time required for a 90 percent reduction of temperature on the linear portion of the cooling curve defined in Eq. 5D2-7 ' superficial mass velocity (pve). height of bed (stack) surface heat transfer coefficient log factor - dimensionless jC log factor at the geometric center jm log factor for the mean temperature in the apple js log factor for the surface temperature thermal conductivity weight of apples in the room Biot number ha/ k - dimensionless pressure difference across the packed bed xiv AP A PT cfm pressure difference across ceiling wall and floor total pressure difference the fan has to overcome quantity of air, cu. ft. /min. heat of respiration variable distance from the center of the apple Reynold number - dimensionless temperature of the apple initial temperature of the apple temperature of cooling media (air) temperature of refrigerant the apparent initial temperature as described by the linear portion of the cooling curve temperature of apple after 24 hours cooling initial temperature of surface surface temperature time average flow velocity in the cross section available for flow thermal diffusivity k/CP-p volume of void volume of bed void fraction viscosity' fluid film viscosity density XV ,4 L1 ([1 *1 0. ’ SE C TION 1 INTRODUCTION The objective of CA (controlled atmosphere) storage is the extension of the storage life of food products beyond that attainable by conventional cold storage practice. At the present time, com- mercial application of CA storage is limited principally to apples and pears, but it is envisioned that this technique will also be used commercially for the storage of other fruits, vegetables and food products. CA has become more complicated and whereas a few years ago there was only one method of producing the atmosphere and con- trolling carbon dioxide, today there are many to choose from. This thesis deals with the CA. storage of apples, but most of the observations are applicable to other types of produce. The objective of this study was to assemble information in an orderly manner that will make possible better decisions in the area of CA construction and Operation. This is an engineering analysis of some of the unit operations available for CA storage and some new approaches and suggestions for improving CA operations. This study was limited to the control of the atmosphere, sealing against gas and vapor transfer and cooling and air distribution in the CA room. A C 11110le 'm‘DTmat gtOrage ‘ availabh pared a! A critical literature review was made of the unit operations involved in the control of the atmOSphere: The necessary design information for the selection and operation of equipment in a CA storage was assembled. The performance and costs of operating the available systems to control the atmosphere in the room were com- pared and analyzed, and specific recommendations were made re- garding which method to use for a given set of conditions. A critical literature analysis was made of the basic problems involved in using a vapor barrier and a gas seal. The problems of gas leakage in the CA storage structure were investigated. The pur- pose was also to discuss and show possible solutions to water vapor transfer, CA storage gas tightness and pressure and temperature cycles in the CA storage structure as influenced by weather conditions, material used and various design procedures. Since the refrigeration requirements for cooling the apple in the storage are large for the first few days of the storage, the possibility of using liquid nitrogen for partial precooling and atmosphere control was analyzed. As an improvement of air distribution in the room during precooling and storage is needed, the objective was also to present a possible new approach to precooling and air distribution design in a room filled with apples. CO ‘1‘: n k l .C‘ I“ e g Cirntr 0.2m" ’ ‘1- '1 A0, SE C TION 2 THE CONTROL OF THE ATMOSPHERE 2A. Introduction After harvest, living fruits carry on respiration, which is slowed down by lowered temperatures, and by reducing the oxygen content and increased concentration of carbon dioxide in the air around the fruit. Controlled atmosphere (CA) storage is a method of keeping living fruit longer by refrigeration, reduced oxygen and higher than normal carbondioxide concentration in gas tight rooms. The optimum re- quirements for the different varieties vary and can only be determined experimentally. The purpose of this section is to analyze the various methods of controlling the atmo Sphere . 2B. Re stric ted Ventilation The composition of the air in a gas-tight storage room containing fruit will be changed due to fruit respiration. The fruit absorbs oxygen from the atmosphere and gives off carbon dioxide in approximately equal volumes. As the concentration of carbon dioxide rises to 5, 10 or 15 percent the concentration of the oxygen will fall to 16, 11, or 6 percent, the sun value 0 carbon if a cor become drops 5 into the the des T1011- of 15 perc COEEEm CO2 an Memo bEIOW I y. . “53111.3 the sum of the CO; and 02 remaining approximately 21 percent, i. e. , the value of the original oxygen content. All the oxygen will be replaced by carbon dioxide unless ventilation is introduced at some intermediate level. If a complete lack of oxygen persists over a period of few days, the fruit becomes alcoholic and develops off flavors. Therefore, when the oxygen drops slightly below the desired level, fresh outside air is introduced into the room. Ventilation can be used to regulate the atmosphere but only when the desired oxygen and carbon dioxide level totals 21 percent. A combina— tion of low oxygen and high carbon dioxide (for example 3 percent 02 and 18 percent C02) will cause damage to the fruit because of the high COZ content. The combination of low C02 and high 02 (for example 5 percent C02 and 16 percent 02) will cause the respiration rate to be high and the deterioration of the fruit to be precipitated. Our present needs are low oxygen and carbon dioxide levels well below the total sum of 21 percent, which cannot be achieved by restricted ventilation only. Even though restricted ventilation operation is very simple and the cheapest one to build, it is not recommended for use today and it will not be described nor analyzed any further. The control of the atmosphere to achieve desired CO; and 02 level is analyzed in the following sections. 21.; cent. 1 oxide! "scrui non u: a—* inj ‘ s. 2C. Removal of Excess Carbon Dioxide 2C1. Introduction Other combinations than CO; and 02 concentrations, totaling 21 per- cent, (restricted ventilation) can be obtained by removing the carbon di- oxide produced by respiration. Removal of C02 can be effected by "scrubbing" the atmosphere from the storage room through an absorp- tion unit. The main requirements for scrubbers are: (l) economical in use, (2) does not give off volatile products likely to damage the stored produce, (3) operate also as an air purifier, (4) can be operated by unskilled labor. The carbon dioxide absorption system of a CA storage room should be designed on the basis of the quantity of fruit in the CA room, the carbon dioxide production rate of this fruit, and the carbon dioxide level in the storage. Each room must be equipped with an absorber of adequate capacity to handle the most difficult carbon dioxide removal job. This condition in Michigan and in the Northeastern United States is produced by the McIntosh variety of apples which requires a storage temperature of 36° to 38°F and a carbon dioxide level of 2. 5 percent during the first month of the storage (Pflug 1960). The CO; solvent should have high CO; solubility which will result in minimum solvent rate needed. The solvent should be relatively mwvde andpref Us confinuc anempt mhmg: U fl.) U: (I) r+ hi non-volatile, cheap, non-corrosive, stable, non-viscous, non-foaming and preferably non-flammable. Usually the apparatus used for contacting a liquid and a gas stream continuously is either a packed tower filled with solid packing material, an empty tower into which the liquid is sprayed, or a plate-type unit con- taining a number of bubble-cap or sieve plates. Ordinarily, the gas and liquid streams are made to flow countercurrently to each other as they pass through the equipment. The physical and chemical principles involved in the scrubbing process and the application of scrubbers in the chemical industry will be discussed. The scrubbers used in the chemical industry are quite different in scale, available source of heat (steam) and operational pressure from those used in CA storages. This comparison is designed to show design possibilities, effect of variables, problems involved and their solution and potential improvement in scrubbers for CA storage use. The application of the scrubber in CA storages, its general design, specific problems, advantages and disadvantages in use will be dis- cussed. The economic analysis is given in Section 2J. The following is a broad division of C02 removal processes con- venient for this discussion: Physical absorption processes Chemical absorption processes (A. (F) Ix) liquids .‘ 50:136 7- Ere Removal using molecular sieves Removal using selective membranes 2C2. Physical absorption processes General The solubility of gasses in liquids is normally expressed in liquids is normally expressed in terms of Henry's law; p = ng (2C2-1) where: p = the partial pressure of the soluble gas the mole fraction of gas in liquid phase xg H Henry constant Henry's law states that the quantity of gas dissolved in a given quan- tity of solvent or solution is directly proportional to its partial pressure over the solution, therefore, the higher the pressure of the gas stream, the higher the partial pressure of the C02 and its solubility in the solvent. H is constant (or practically constant) for a given tempera- ture, but increases with an increase in temperature. Absorption in Water The solubility of carbon dioxide in water is relatively low. The quantity of C02 absorbed is practically directly proportional to its partial pressure and decreases with increasing water temperature (Perry 1963). The data of the gas-liquid equilibrium relationship are used to determine the quantity of water needed to absorb the required carbon dioxide, Ex- tensive work has been done on physical absorption processes of C02 in 7C2- 11021 ‘E a 6 ab .9 -\A II.— water in the chemical industry. A summary of mass transfer rate refer- ences for different types of packing and tower sizes is given in Table 2C2-1. Yoshida (1958) summarized the work done in the area of absorp- tion of CO; in water and methanol in packed absorption columns. He studied the absorbance of pure CD; into water and methanol in a column packed with spheres, Rashing rings, and Berl saddles as well as in a bead column containing Spheres. Reasonable correlations were obtained. Sherwood and Holloway (1940) have correlated their data on the desorption of oxygen, CO; and H2 from water in a packed column using the equation kLa:a D _L_ p (2C2-2) 0.25 0.50 H ) pD where: La liquid film coefficient, lb. mols/ (hr. ) (cu. ft.) (lb. mol/cu. ft.) D diffusivity of solute gas in liquid, sq. cm/ sec. 0 empirical constant given for different packing materials L liquid rate, lbs. /(hr. ) (sq. ft. tower cross section. ) 11 liquid viscosity, poises p liquid density, lb. /cu. ft. 3.2 883.8% om 8:... 338m 833 Noo 82 33383 m. .3 888m 833 Noo $2 38383 on 8838. 38m 833 N06 32 33.33 m 8838 m2me 833 N00 32 33803 m .3 8833 m2me 833 Noo ova poozionm om mwgh chmmm .8pr N.OU $2 85000 opmsvm om 933mm.“ 33m 33m? NOD $2 .3 8 «.86 o 35.3 338$ 833 Now N2: 38338 3 .o 38.3 333m 833 Now $2 80M 3 .8 3:8 328$ 833 Noo 3.2 3330.3 m .3 8:8 333m .833 Noo 32 «fish m. 8:: 938$ .833 N00 mocopomom .5 690. wcfiomm Eofiom 330m 3368me .8308 moocohmmom mo pmwAlmpmnnomnm poxoma 5 Ban sommcwfi mmmz .auNUN MJmAwE "WP LC..... 10": II\ 5'1 ) {—1 .. V“ 10 The important results from their studies are: 1. Liquid film coefficient k increases rapidly with temperature, La presumably due to a decrease in liquid viscosity. Data for carbon di- oxide and oxygen are practically identical since the diffusivity of CO; 0.023T as e where T-water La and Oz in water differ only by 15 percent. k temperature °C. 2. Effect of gas velocity-k a is independent of gas velocity in the L normal velocity ranges. 3. There is a marked effect of liquid rate on kLa' 4. An addition of wetting agents to the liquid or coating of the packing with paraffin caused a considerable reduction of kLa' 5. The data showed relatively small variation of kLa with size and type of packing, liquid distribution at the top of the tower and man- ner of placing the packing in the tower. Cooperetgl. (1941) suggested that in a COz-water absorption tower, the ratio of the liquid flow to gas flow must be very large be- cause of low solubility of the gas. A scrubbing apparatus particularly suited for this application using cascade packing is described by Cooper (19451 The CO; absorption by water and water-glycerol mixtures was also made using bubble-tray absorption columns (Walter and Sherwood 1941). It was shown that the higher the viscosity the lower the plate efficiency. There is no advantage in using water-glycerol mixtures nor bubble-tray absorption column in CA room. 11 The application of the water scrubbirgjn the CA Palmer (1959) was the first to suggest the use of water absorption- desorption system to remove carbon dioxide from CA storages. Several types of water scrubbers are in use in New York State (Smock 1960, Kendenburg 1959) in which the carbon dioxide is absorbed in the water circulated over a wet type refrigeration evaporator and removed from this water in an aeration tower located outside the CA storage (Fig. 2C2-1). The disadvantage of the combination of this water scrubber and refrigeration coils is that when operating at low tem- peratures (31- 32°F) a wet coil would require an antifreeze material in the water, which significantly reduces the carbon dioxide absorp- tion efficiency of the system. This disadvantage and the possible use of dry type refrigeration evaporators led Pflug (1961) to design a single tower water absorption- desorption system to remove carbon dioxide from CA storage. This system operates independently of the refrigeration system and its design is based on chemical engineering principles of absorption- desorption packed-bed tower. (Sherwood and Pigford 1952, Leva 1953). Starting at the top of the tower, (Fig. 2C2-2) water flows downward through the absorption section where it comes into contact with air from the storage, which may contain, for example, 5.0 percent carbon dioxide. Some of the carbon dioxide diffuses into and is absorbed by the water. The water next flows through the desorbing section where it comes into contact with normal atmosphere WALL "l 1 ' I ‘1‘ FULL . @ 1" JET NOZZLE I , 3 4‘ .. ’1 IO VALVE—*t ‘f H 1 _. l _ _ _ _ _ __ ___ 1 10 .L "--------- ‘ q . ,9 VALVE / SPRAYS é” ‘1 .. / st ’ Co 0 4 COIL SECTION ,3 .. v- t 1 K / ‘ t Z : _—_—_—_—_ / 1 _WTEF_ _ J , n ‘aSE OF 3. ‘ _—_—_ ‘ 4— BLOWER g ‘ - 33w , DRAIN—h2g5 .1. Fig. 2(‘2-1. Simple basw (1(‘SlgIIOI‘IOH't’I‘ \uth I'ull jot spray n07.211- for water scrubbing (Smm‘k £1 §1_1 1950; 13 V' . ------- . ----- \ 11 / c(>~\ as | I \r l A G \ A \\ I I l' “V l : . f . 11 '1 \ l. Q) ABSORBER AIR . 0 BLOWER ABSORBER 1 ¢,-___/ SECTION f ’ 1 1' EEEE E f}. . \ f \ .' \\ 11‘3”; \ >‘\| II‘ t \\ .Jl L“; “V 0 TRAP ‘ 1‘ i 0 ‘I‘ t :1 | ; \{N ‘1, E \V I \\ 1} II I C! I \\ OESORBER AIR BLOWER \ +115 -------- - {I CONTROLLED 1 {I ATMOSPHERE g3. ROOM 3 DESORBER 1} g SECTION {1 : {I \\ {I ' 1., 11$ S'\ t "9’; J \\‘; _ \\ II I; q. I F; ~ I, '5 \\ :t— ’ :5 \ ‘ PUMP * , -- u-«q ‘1 3; \\ 1 l -x 1'33, \_‘\\1 KTo DRAIN Fig. zcz-z. Cross section diagram of water absorber for CA fruit storage (Pflug 1961) 14 that may have but 0. 03 percent carbon dioxide and where some of the carbon dioxide is desorbed from the water and diffuses out of the water into this low carbon dioxide atmosphere. The water normally is recirculated for continuation of the absorption- desorption process. The mass transfer of carbon dioxide to or from a water solution is liquid-film controlled and can be described by the equation: : A - N KLa Vt C (2C2 3) where: N lb. mol of solute transferred per hr. lb. mols KLa overall coefficient cu. ft' hr. lb. mols of solute cu. ft./soln. Vt tower volume, cu. ft. lb. mol of solute cu. ft. /soln AC log mean concentration difference The diffusion coefficient, liquid viscosity, temperature and liquid flow rate determine the overall coefficient KLa for carbon dioxide in water. Liquid flow rate is the most important of these variables; the overall coefficient varies approximately directly with liquid flow rate. Both the diffusion coefficient and viscosity are temperature dependent, increasing the temperature increases the diffusion‘coefficient and re- duces viscosity. Therefore, K is increased by an increase in the La 15 diffusion and decrease in the liquid viscosity. Under usual conditions, gas flow rate does not significantly affect the absorption coefficient. The carbon dioxide concentration difference at any point in a liquid-film controlled system is the difference between the carbon di- oxide concentration in the bulk of the liquid and the equilibrium liquid concentration of the carbon dioxide in the gas phase expressed as molar concentration. The concentration difference of the system is the log mean Of the inlet and outlet conditions. Since carbon dioxide- water equilibriums are the function Of temperature the concentration difference is a function of temperature; increasing the temperature decreases the equilibrium concentration. The equilibrium concentra- tion (expressed as mole fraction) Of 5.0 percent carbon dioxide gas in water is 3. 08 ><1o'5 at 77°F (25°C) and 4. 88 10'5 at 50°F (10°C). Therefore, the change in absorption rate due to a change in tempera- ture is quite small since the increase in K a brought about by an in- L crease in temperature is Offset by the decrease in concentration difference. In general the carbon dioxide level of the CA storage is the largest single factor in determining the carbon dioxide concentra- tion difference, and therefore, the carbon dioxide removal rate for a specific system. The quantity of carbon dioxide removed from a CA storage in- creases linearly with packing depth, however, 4. 5 ft. was mentioned by Pflug as the approximate maximum economical packing depth for 16 both the absorption and desorption sections. This was based on gas pump- ing requirements; up to 4. 5 ft. a centrifugal blower can be used, above 4. 5 ft. some type of positive air pump is required. A detailed descrip- tion of the MSU water scrubber is given by Pflug (1961) (Fig. 2C2-2) but the most important and unique is the liquid trap section which enables the construction of absorption and desorption section in one compact unit. In the liquid trap located between the absorber and desorber sections, water flows from the absorber section to the desorber section without removal of CA storage room gas from the absorber section, or the addition Of outside air from the desorber section into the absorber section (Fig. 4A-1). The MSU water absorber was designed and con- structed so that caustic soda could be used in an emergency. In the same paper Pflug (1961) also showed a way to size the MSU water absorber needed for specific application. Advantages of water scrubbing 1. Compared to caustic soda scrubbing, water scrubbing is more economical because it saves on the use of caustic soda. 2. It is cleaner and less caustic in operation than the caustic soda scrubber. 3. The manipulation is faster with water scrubbing as caustic soda need not be drained out. 17 4. Prevent the continuous increase of the ethylene concentration and reduce the ethylene and the nonethylenic volatile concentration in the CA room. Disadvantages of water scrubbing Oxygen is slightly soluble in water and some oxygen is added to the CA room with water aeration in addition to the oxygen added in the air that replaces the carbon dioxide. An analysis Of the effect of water absorption of carbon dioxide on the operation of the CA storage room (Pflug 1960) showed that tight CA rooms filled with fruit should require but two or three days longer tO develop a 3 percent oxygen atmosphere than similar rooms operated with caustic soda absorber. This analysis also showed that rooms which are borderline in tightness or partially filled will not operate satisfactory when a water absorption system is used for carbon dioxide removal. Because of the reasons mentioned above the water scrubber is not recommended for use when the de- sired COZ level in the room is less than 5 percent. Other Physical Absorption Systems In order to have a complete picture of physical absorption of C02 in the chemical industry two other processes will be mentioned: 1. Absorption is methanol. The solubility of C02 in methanol in- creases sharply with a decrease in temperature (Hoogendoorn 1963). It is practical in the chemical industry because of the advantage of the absorp not pr; :nd we r»! ried 0 I‘EVEr the ‘50 Sirip; SOFpti 18 absorption of other gases like hydrocarbons, hydrogen sulphide but is not practical for CA because the methanol is more expensive, volatile and would introduce extra refrigeration expense. 2. The Fluor Solvent process (Hoogendoorn 1963). In this process an organic solvent such as propylene carbonate is used. This process, as in all physical absorption systems, is most suitable with high total gas pressures and high concentration Of carbon dioxide. Since there is no high gas pressure and high COZ concentration it is not suitable for use in CA storage. 2C3. Chemical absorption processes General Absorption of gases with low solubility in water is frequently car- ried out by using liquid absorbents that react either irreversibly, or reversibly-forming a compound that is easily decomposed by heating the solution. In the last case usually the solvent is recovered in a stripping operation using heat or air. 8 There is no sharp line dividing pure physical absorption from ab- sorption accompanied by rapid chemical reaction. Most cases fall in the intermediate range. Table 2C3-1 lists experimental studies on simultaneous absorption (or desorption) and reaction systems men- tioned in the literature. The use of experimental values is particularly important when absorption or desorption occurs simultaneously with LL33)? 11.1.1'1 I. Illllll‘ll 11..-.IIS'III’ I!" I I ‘I .II «11". 9.1—! ‘0 I. u 1.. - - u 1 I .1 a u A n u v 1 .— l- .U V. U AV uns.~ Ila“: IVA-ltihvnun-dh .Urivu‘piu... nhflv“ V n h ~ M h. r n: . . . A. s \ . V.‘ V nh‘ I‘. N I V~£F~U QS W‘xQ E ~ * V ‘ ‘\ ~ I v at h I. v Ch . u . . ~ ' O l9 mflwaw HOCmfimw mmod .3308 w mcfimwp H005 0002 .8wa ~00 0583 3:350 3mm; knowosw NH magnum 0:02 033.3 300 0583 Hoswfio 332 02803 3 .3 982 833 300 0598 3:050 33 3 Box 832 833 300 332 883380 3 3533333 m00 332 833 N00 mmfippmm Comm 333 3338.3 NH 333 3333303 300 332 833 300 ~32 83m 3 .3 .3 3.3333333 300 3.32 .833 300 332 803380 3 3533333 300 3M 833 N00 33$ 85 333 3888800 300 NM 3m 833 N00 332 838m 3 .3 353.3333 300 3M 3m 833 ~00 mm? 0003:0030 w 0953 wcwmmod $03.2 3033.3 300 332 388m om 33:8 :30 $032 833 300 332 384 3 8:333 U833:: $032 833 300 332 384 3 8:833 23m 3532 833 300 332 83333 3.532 833 N00 3 32 38.3 3 3.333333 $0.32 833 N00 332 82m 3 .3 .3 38.53333 3532 833 ~00 33$ 3334 3 m 2 3838333.. 3 33333.33 33032 833 3o o 33.3 83m 3 .3 .3 3583m mom 833 N00 .5 093 mcwxowm Eowmom ”EOE/How 8.30m 930835 30308 mongoose?!“ .Ho «mddlmponhomnm @0393 0030309 Hwoflbono 5 33.3 30.30.95 00.32 L Aim UN HJMNa2CO3 + H20 (2C3-1). 11. C0; + N32 CO3 + H20 :: ZNaHCO3 (ZC3—2). As can be shown from these equa- tions, 1 lb. of pure sodium hydroxide will absorb 1. 1 lb. or carbon dioxide. One- half is absorbed in each step, but step I takes place much faster than II. Reaction II is reversible, sodium hydroxide will react with sodium bicarbonate to form sodium carbonate. The rate of absorption of C02 is affected by the state of comple- tion of the chemical reaction. At the beginning, with fresh NaOH, there is great affinity of NaOH for CO; (Pflug el al. 1957a). The rate of absorption decreases due to the increased concentration of sodium carbonate as reaction I nears completion and decreases even more as the sodium carbonate is neutralized to sodium bicarbonate. Blum _e_t _a_.l. (1952) developed an equation for representing the C02 absorption by hydroxide in packed tower. 0.84 (OH—NCO?) Na:0.0176.LIn h (“ml-09 (2C3-3) flare Na Sflrpdo absorb rx) 23 where Na moles transferred, 1b. mols/ (hr. )(sq. ft.) of tower cross area \ .. LIn liquid rate, lb. mols/(hr. )(sq. ft) (OH’) arithmetic average (top and bottom of column) normality of hydroxide. (CO? arithmetic average (top and bottom of column) normality of carbonate. H' ionic strength h height of packing, ft. Several researchers (Table 2C 3-1) have extensively studied the absorp- tion of sodium hydroxide in absorption tower in the chemical industry. Pfluggtal. (1957c) used the chemical engineering principles dis- cussed by these researchers and presented the fundamental of C02 ab- sorption with NaOH as applied to CA and designed the MSU packed tower absorber. In addition to the effect of state of completion of the chemical reaction mentioned earlier, Pflug studied the effects of liquid flow, ab- sorbing surface area, liquid temperature, concentration of CO; in the gas atmosphere and the air flow rate. His results are summarized as follows: 1. The rate of absorption increases as the rate of solution flow thrOugh the absorber increases. 2. The rate of absorption is proportional to the surface contact area 0f the solution with the atmosphere containing C02. sorbi: tion. {'16 u ‘90 ir . :9? o forai firm 24 3. The rate of absorption increases as the temperature of the ab- sorbing solution is raised. 4. The rate of absorption increases with increase in C02 concentra- tion. 5. The rate of air flow through the absorber does not materially affect the absorption coefficient. High air flow rates are undesirable since they waste power and cool the absorbing solution. There are three types of absorbing systems that were commonly used in the U. S. : The barrel type designed by Smock and Van Doren (1941) (Fig. 2C 3-1), the brine spray type that was first suggested by Smock and the system is described by Kedenberg (1953) (Fig. 2C3-2) and the MSU packed tower designed by Pfluggtal. (1957c) (Fig. 2C3- 3). The barrel type absorber (Fig. 2C 3-1) uses perforated plates in the upper portion of the tank while the lower portion is used as a reser- voir. The NaOH solution is pumped from the bottom reservoir to the top of the absorption section and allowed to flow down through the per- foration of the plates. The air from the storage room is forced up through the holes in the plates and back into the storage room. The brine spray diffuser (Fig. 2C 3-2) is a standard refrigeration equipment. It serves the dual purpose of evaporator and absorption tower by replacing brine with NaOH solution. Absorption takes place when the NaOH solution is sprayed on the evaporator pipes over which air is rapidly circulated. 2.5 The MSU packed tower (Fig. 2C3- 3) operates independently of the re- frigeration system of the storage room. It consists of an absorption tower, reservoir tank, circulating pump, air blower and connecting piping. The atmosphere of the CA room is forced by an air-circulating blower through the packed absorption tower (Interlox saddle packing is recommended). It enters the tower beneath the packing at a point just above the liquid level of the filled reservoir and returns to the room from the top of the tower. The liquid is pumped from the bottom of the reservoir into the top of the tower, where it flows over the packing in contact with the room atmosphere and drains into the reservoir. Angelini (1956) observed that the time required to utilize 90 percent of the NaOH was more than 60 min. /lb. for barrel unit, 12 to 22. 5 min. /lb. for brine-spray units, and 18 min. /1b. for the packed tower. Eckert_e__ta_l_. (1958) showed in their studies using 4 percent NaOH solution for the absorption of C02 in the chemical industry, an improve- ment of 30 up to 67 percent in the K a using the new Pall rings compared G to the rashing type. Advantages of hydroxide solution scrubbirg. This is a very efficient C02 scrubbing operation which permits reduc- tion of the oxygen level in the room in a short time. The use of hydroxide solutions is advantageous at the first period and in emergency in some water scrubbing units. J 3.02 E:0m:....; :c: :0: W.\\\\\\\\\\k§ -9202: 20.2? 0:2: 0-3.5. 31.; 333 2.2331 s 32.88 Y 352533 m. 35528 \ MU N v... z. \ 3233 «2]? a \ mono: J n H H H H x x. O o a \ \x o o o o o WI:OU\ o o o o o \ WWQ_Q TEEESQILT . . . . . 54339333 0“. C . \ x L 3:82AM 1% EEE rdmaw \ a m \ by O \ .3; 283 So 23 \ 033.83 «2‘ 3mm: _::;m:<.. 2:: :2. . 1.200;: .03.: TE .3: o3; 2:25 m_o>mmmmm 283 r. woaxozl 201... IV! 32 II I l .I l Iluvvh» I H lllllllllllin. .0222 lllllll 1 3:03 33:4,3 IIIIIIIIII 823853 I l llluIIHH H ”[L _ 2003 I 334323 [II 3.. E 32 .7703 .32 0230 «0:03 mmnjo .633 mono... I O ‘ I .w...... . . .... H .-« ...—. - “-1... . .. .- a- nu. -—..~-L.-La-n _,,m,.-, . '—- Q . -fl —-.4_-..-—-.......--‘ 27 5.2: mm .mm mar: 33%... 5:28.? pm: a: mo 883% 83838 .m-mu~ mE 38848.5 gs O val—(kg L fl All \l 1“ ‘ 28 Disadvantages of hydroxide solution scrubbing The CO; reacts irreversibly with the NaOH causing a decrease in rate of absorption during the absorption period and frequent recharging with fresh NaOH solution. This last step is expensive, troublesome, laborious, and is a danger to health. As the solution is highly corrosive there are problems with the construction of the absorber, charging the absorber and discarding of the spent liquor. Dry Lime Scrubbing Eaves (1964) observed that fresh hydrated lime Ca(OH)2 will absorb carbon dioxide from a CA storage room atmosphere when the hydrated lime is itself still inside closed paper bags. The chemical reaction is: Ca(OH)2 + COz—-)Ca (:03 4 H20 Theoretically, one lb. of lime will absorb 4. 4 cu. ft. of carbon dioxide. Since one bu. of McIntosh apples at 38°F will produce approximately 4. 7 cu. ft. of carbon dioxide during a 6-month storage period, the dry lime requirement would be approximately one lb. per bushel. Eaves (1964) estimated that 1-1. 5 lb lime was being used per bu. for apple storage in Nova Scotia; and indicated, based on Smock's findings, that as much as 4 lb. of lime per bu. was used in New York State. Usually no tests are made to determine'whether the lime has absorbed 29 100 percent of the carbon dioxide it is capable of absorbing, therefore sometimes the lime is replaced though considerable carbon dioxide removal capacity remains. There is also the fact that lime may not be fresh and that it will have absorbed a considerable amount of carbon dioxide from the atmosphere before it is used in the storage. Observing the chemical reaction above, we note that water is pro- duced by the reaction; however, it is also possible that the dry lime absorbs some moisture from the storage room atmosphere. The calcium hydroxide reaction with carbon dioxide produces 103 Btu's per mole; this heat has to be taken into account when calculating the total refrigeration load. Normally the dry lime used for carbon dioxide removal contains 90 percent calcium oxide and 10 percent magnesium oxide and is of 325 mesh. Eaves (1964) reported on a survey to evaluate quality characteris- tics of fruit that had been stored in CA storages where carbon dioxide was removed using dry lime. He reported that there was no weight loss or odor due to the dry lime and that 90-92 percent relative hu- midity was easily maintained in the storage. The operation of CA storage rooms when dry lime is used to remove carbon dioxide is some- what simplified in that the carbon dioxide level normally remains quite constant due to the rather uniform removal of carbon dioxide by the dry lime. Oxygen level problems are somewhat reduced since there is no oxygen brought into the room during the scrubbing operation as 3O happens when a water type scrubber is used. The dry lime can be used for very low carbon dioxide levels, for example, in the Pacific Northwest the standard practice is to store Delicious apples with a maximum carbon dioxide level of 2 percent. It is relatively difficult to maintain this level with a water type absorption system; however, the dry lime functions very well in this situation. The movement of carbon dioxide from the storage room atmosphere into the dry lime will take place according to the laws of diffusion. The higher the concentration of carbon dioxide in the storage room atmosphere, the faster will the carbon dioxide move into the lime. When the lime remains in the bag diffusion into this mass of lime must be considered. It is probable that when lime is discarded, the outer portions of the bag are more fully saturated with carbon dioxide than the inner portions. Also, when high carbon dioxide levels are used, it is possible to utilize the dry lime more fully than when low carbon dioxide levels . are used. The larger the dry lime room, the greater will be the ability to utilize the total capacity of the bags of the lime. There is considerable difference among different sources of lime. It is suggested that some type of absorption test be run on the lime to determine its carbon dioxide absorbing ability or more preferably, the lime should be certified from the manufacturer indicating that it is fresh and that it meets carbon dioxide absorption standards. The permeability of the bag will effect the dry lime absorber performance, especially where the bags are not open or holes are not punched in the bag. Plastic bags 31 must not be used. A diagram of the dry lime scrubber is shown in Fig. 2C 3-4. The dry lime box has to be air tight since it is an integral part of the CA storage room. The dry lime box should be designed in such a way that the dry lime can be moved into and the spent lime removed out of the lime box rapidly and with a minimum amount of labor. Advantages of dry lime scrubbigg. l. The operation of the room is simplified in that constant levels of carbon dioxide are easily maintained and rapid oxygen reduction rates are obtained. 2. Power requirements are low and the only part to deteriorate or cause problems is the blower and damper system. 3. High carbon dioxide removal capacities are available during the pull down period on a no penalty basis. 4. Operation is very simple. 5. There are no corrosion problems. 6. The system is relatively easy to construct with local labor. Disadvantages of dry lime scrubbing. 1. Relative humidity in dry lime storages may be slightly lower than in storages using a water absorption system. Also, the water absorption system may have greater odor removal capacity than the dry lime system. 2. The dry lime itself creates a number of problems; the sources :.::_ 7:31.47 .r...m:.:;.r. L31: 6.; .I: ._;.::.:.r. 2:: .2: .T.. v: 9:3 @293: .ooo o>_o> , I 4! ECON— <0 3:960 .222 33 of supply are limited; the variability in supply is large; the material is bulky to handle and to store; the dry lime presents a real disposal problem. 3. The dry lime room requires a considerable amount of Space; often times space that could for the same cost be utilized as fruit storage space. A practice is gaining use which is to put approximately 1/10 lb. of dry lime per bu. of apples inside the CA storage room and there- fore the dry lime box can be made smaller. This lime is put either on top or in some obscure place where it does not normally take up space that would be occupied by apples, for example, underneath the floor mounted diffuser, etc. This dry lime, of course, has an initial high affinity for carbon dioxide which means that it aids in controlling the carbon dioxide level during the pull down period and since no oxygen is added as carbon dioxide is removed, this aids in oxygen come down. The dry lime tapers off toward the end of the season as it becomes saturated and at the point where the dry lime ceases to be very im- portant the water absorber can easily carry the carbon dioxide removal load. Reversible Chemical Absorjgtion Carbonate Solution When carbon dioxide is dissolved in an aqueous solution of sodium or 34 potassium carbonate the following reversible reaction occurs: N32 CO3 + C02 + H202 2N3HCO3 (2C3‘5) The solubility of C02 in such a solution depends on the ratio of carbonate to bicarbonate, the total amount of carbonate in the solution, the tem- perature, and the partial pressure of carbon dioxide in the gas. Harte _§_t_e_l_l. (1933) expressed the relation in an empirical formula, which reduces to (137sz )1'29 P : N8. (2C3'6) COZ S(l-f) (365-T) where P partial pressure of C02, mm H COZ f fraction of total base present as bicarbonate NNa sodium normality S solubility of CO; in water under a pressure of 1 atm. of C02, g. mol/ liter T temperature °F Eq. 2C3-6 has been tested over the temperature range of 65° to 150°F and sodium normalities from O. 5 to 2.0. Sherwood and Pigford (1952) reported that the data on the potassium system may be approxi- mated by the equation: (45fZNK)1'29 co2 = S(l-f) (302-T) (“334) P (‘1 In (D I mm (D 35 where NK is the potassium normality. Eq. ZC3-7 has been tested over the temperature range of 30° to 100°F and potassium noramlities from 1.0 to 2.0. Crystals of bicarbonate can be formed above a concentration of 30 percent carbonate equiva- lent. Use of hot potassium carbonate There is no advantage in using carbonate compared to water at room temperature. The solubility of C02 in the carbonate is appreciably high at higher temperatures, therefore the hot potassium carbonate process was deve10ped. It was developed especially for bulk removal of C02 from a gas when a high degree of removal is go_t necessary (down to O. 5 percent) and where the inlet concentration of C02 is in the range of 5 to 50 percent. Sieve trays have been found advantageous for use in CO; absorption and regeneration in hot K2 CO3 solution, a substantial quantity of per- formance data have been reported by Buck and Leitch (1958). For this system, sieve trays are reported to be more economical than packed absorption towers. Recently small quantities of various chemicals have been added to the carbonate solution to accelerate the absorption process. The Giammorco-Vetrocoke process (Hoogendoorn 196 3), is a modification 01 13.6 sorp 36 of the hot potassium carbonate process. In this process the C02 ab- sorption rate is considerably increased by the addition of selenious, telluric, arsenous oxides and amino acids to the potassium carbonate solution. The kinetics of this pseudo first order reaction was studied by Richards (1964). The use of these additives will not increase the maximum volume of carbon dioxide which can be absorbed under equilibrium condition. However, it does make a closer approach to equilibrium possible, resulting in more COZ actually absorbed per volume of solvent and also smaller towers can be used. In the re- generation step, the same effect produces a leaner solvent with less residual COZ, which means that a lower outlet concentration of CO; in the scrubbed gas can be obtained and also the volume of CO; that can be absorbed per volume lean solvent is larger. The usual absorption tem- perature (Hoogendoorn 196 3) is 140°F and regeneration takes place at 220°F. Air stripping is also possible (at 160°F). An advantage of potash (KOH) solution containing arsenous oxide is that the arsenous oxide acts as a corrosion inhibitor so that normal carbon steel can be used in constructing the equipment. Mayland (1964) described also a simi- lar process in which ethanolamine and corrosion inhibitors such as di- chromate are added to the hot carbonate solution. Fundamental investi- gation on the effect of additives by Ryan and Carter as described by Jeffreys and Bull (1964) reported that the absorption rate was increased by small addition of methanol, many sugars and particularly or (I) 37 formaldehyde. The explanation was that the additives depressed the sur- face tension of the absorbent which in turn brought about an increase in the absorption rate. Jeffreys and Bull (1964) also showed that the ad- dition of glycine to sodium carbonate solution increased the rate of absorption. This is believed to be due to glycinate ion accumulating in the surface of the liquid catalyzing the absorption reaction. The application of carbonate scrubbing in the CA The Sulzer Company in Switzerland (Meyer 1961) is using potas- sium carbonate solution in its COZ absorber. They claim the ability to wash out carbon dioxide economically in a circulation system prac- tically without pressure and without thermal requirement. The choice of this process was decided mainly because of the following factors: K2 CO3 is odorless, inexpensive, non- toxic and as an absorption medium it can be regenerated with air. It is possible also to install the absorp- tion unit even at ambient temperature of 32-50°F, without adversely affecting either the economical operation of the absorber or the re- frigeration condition in the CA storage. The Sulzer Company mentioned the use of a new type of mixing nozzle for distribution of liquid in spray towers, drop washers, washing towers with filter elements, sprinkler towers or packed washer, that doubles or triples the absorption rate compared with other systems. The principle of the gas bubble washer is described in a schematic diagram (Fig. 2C3- 5). The mixture of air and C02 1 plates, it the men mixture . mixture at the to] Which se in The 1i: pipe, ai‘ air mixt 30p of & 3050mm. mixture. prOCQSE ’he SYS‘ ‘3ng t1 EUTOPE for “Us ~ I. a: 38 and C02 is led via a blower into several mixing nozzles located between two plates, in the bottom of the washing tower. A circulation pump supplies the amount of liquid required for distributing the gas. After an intimate mixture of gas and liquid has been achieved in the mixing nozzles, the mixture rises inside the washing tower. The gas and liquid are separated at the top end of the tower. The liquid spills over into a collecting chamber, which serves at the same time as a separator for the gas bubbles entrained in the liquid. The liquid is led back to the circulation pump through a fall pipe, after which the cycle begins again. After separating the gas and air mixture from the washing fluid, the former leaves the washer at the top of the tower. The process described can be carried out both as an absorption process,i. e. , washing the carbon dioxide out of the gaseous mixture with simultaneous charging of washing fluid, and as an exsorption process, i. e. , regeneration of the washing fluid by blowing air through the system accompanied by enrichment of the latter with carbon dioxide, using the same equipment in either case. This scrubber is used in Europe and was claimed to operate satisfactorily, it can be used also for washing the air in storage with water. There is not any disadvantage in its use from operational reasons. Although water has a better ab- sorption coefficient for CO; than the cold KZCO3 there may be some engineering design advantages in this unit. 39 Blower for air laden with (‘02 ('ircululing pump for absorbent Nnul ' assembly Mixlurc mcrfln“ ‘WIJI‘ Fig. 203-5. Diagram of the Sulzer scrubber (Meyer 1961) 4O Ethanolamine Solutions Aqueous solutions of weak organic bases such as mono-, di-, or triethanolamines can be used to absorb acid gas like CO; and the alkaline absorbent can be regenerated by heating. The vapor pres- sures of CO; from mono-, di:, and triethanolamine solutions of vari- ous concentrations and at temperatures of 0° to 75°C are given in Perry (1963). Amine carbonate is formed according to the overall reaction ZRNHZ + COZ + H20 :: (RNHZ)2H2CO3 (2C3-8) The amines are relatively non-volatile and have a high absorption capacity. The use of monoethanolamine (MEA) as the alkaline ab- sorbent showed greater absorption coefficient compared to the di- and tri- while the solubilities in the different solutions are not greatly different. Rates of absorption of CO; in absorption towers were extensively studied (Table 2C 3-1) and the factors affecting the rate of absorption are mainly: 1. Total amine concentration a0 (moles/vol) and the carbonation £4112; ac (mole COZ/ mole amine) (a) as long as are <0. 5, second order fast chemical absorption takes place (Astarita _€_I._a_1.. 1964) and the overall reaction is coZ + ZRNHZ +RNHCOO" + RNH,+ (2C3-9) Ana 41 An approximate equation forycalculation of RNHZ concentration is (RNHZ) = a0(1-Zac) (ZC3-10) where (RNHZ) is the RNHZ concentration, mols/volume. Absorption rates increase with increasing a0 and decreasing ac , as long as viscosity effect does not influence. (b) When 01C > O. 5, pseudo first- order slow chemical absorption takes place, the overall reaction is RNHCOO_ + coz+ ZHZO » RNH3++ 2Hco3 (2C3-11) Absorption rates decrease with increasing a0 and with increasing ac The step COZ + HZO —> HCO3— + H+ is a slow reaction which is cata- lyzed by arsenite ions and therefore arsenite may help in the absorp- tion of CO; by MEA. Clarke (1964) explained the kinetics of CO; absorption accompanied by chemical reaction but at short contact time. 2. Gas rate. In a packed tower the rate of absorption is sub- stantially independent of the gas rate (G) but when using bubble plate tower at low gas velocities in an atmospheric-pressure COZ absorber (Kohl 1956), appreciably higher efficiencies were obtained. (Bubbling mechanism on the plate is greatly influenced by gas rate.) 3. L_iguid rate. Rate of absorption increased with increase of the liquid rate. 4. CO; concentration. The rate decrease linearly with in- crease of C02 concentration in a solution at atmospheric pressure rel . .1 4'1; .1 phi. range 1 l‘ i. GI 42 and relatively low CO; partial pressure (0 to O. 5 atm). 5. CO; partial pressure. KGa appears to vary with e‘ 3'49 in the range 0->O.5 atm (CO; partial pressure). 6. Temperature. In the range 77° to 167°F, K a increased with G increased temperature up to 122°F and leveled off or decreased in the range from 122° to 167°F. 7. Type of packing. The performance of l-in. Berl saddles is almost identical with that of 1-in. steel Rashing rings but the poly- ethylene Tellerettes (Teller 1958) achieved 23 to 72 percent more ef- fectiveness in mass transfer rate as a function of the. flow condition. Kohl (1956) summarized the factors affecting the absorption of CO; in aqueous solutions of MBA in packed tower in the equation: _ 0.56 0.0067T-3.4p KGa—p0.68[1+ 5.7(o.5 aC)M e ] (2C3 12) where: KGa overall coefficient lb. mols/ (hr. )(cu. ft.) (atm) p. viscosity, centipoises QC concentration of CO; in the solution, mols/mols MEA M amine concentration of the solution (g. mols/liter) T temperature °va ‘ ~ p partial pressure, atm. The annlica In cons age it was : injury by e: *lie ethar eianolami ranging the absorbing arranging place at th J- anc Their abs< illh the a) fer the ( means of 43 The application of ethanolamine scrubber in the CA In considering the use of ethanolamine in scrubbers for CA stor- age it was shown by Mann (1958) that the apples suffered no apparent injury by exposures to vapors of ethanolamines. The most economical of the ethanolamines was the monoethanolamine (MEA.) although tri- ethanolamine (TEA) is used too. Mann suggested two methods of ar- ranging the scrubber unit; (1) having two identical units, one in use absorbing carbon dioxide while the other is being regenerated or (2) arranging some system whereby absorption and regeneration take place at the same time and continuously. J. and E. Hall Ltd..- (1957) have developed continuous absorber. Their absorber consists of two tahks located one above the other, with the absorption taking place in the upper tank. The atmosphere from the chamber being scrubbed is drawn into the upper tank by means of a small compressor and the gases bubble up through the solution, after which the atmosphere containing a reduced concentra- tion of CO; is returned to the storage chamber. The solution of TEA from the upper tank flows down through a heat exchanger into the lower tank where it is reactivated by warming with electric heaters. The hot TEA from the lower tank is returned to the upper tank by a pump through the heat exchanger, and passes through an aftercooler, where the liquid is cooled to approximately atmospheric temperature. The MBA scrubber is very effective in keeping any desired CO; .i . 1. IQ . i. 9 . b. inn 1 .1 9 .u. A19 1F)“ A a‘» O A... .. 9i. like nu... . .nlt s 44 level. It is very useful where the treated gas must be virtually free of CO; but this is not needed in CA storage Operations. A disadvan- tage using this system is that the gas atmosphere is heated unless there is cooling of the solution after reactivation. In the construction of scrubbing units using ethanolamine solutions nonferrous metals should not be used. As there are no operational problems using ethanolamine unit, it will be judged economically only (Sec. ZJ). 2C4. Removal by use of molecular sieves Synthetic Molecular Sieves (MS) are crystalline metal alumino- silicates that have been activated for adsorption by removing their water of hydration. Because little or no change in structure occurs during this dehydration, unusual highly porous adsorbents are formed that have strong affinity for water and certain gases and liquids. MS act as physical adsorbents in which the adsorbed molecule is held within the crystal structure by relatively weak Van der Waals force. When the sieve is desorbed, the crystal retains its original state. Adsorption by MS is characterized by a Langmuir- type iso- therm--the amount of a compound that is adsorbed increases rapidly up to a saturation value; this saturation or equilibrium value usually corresponds to completely filling the internal void volume of the MS. The desorption of MS shows no measureable hysteresis, that is, the adsorption and desorption are reversible. Not only will MS separate rnole eren' of n1 polar c. ithi and ' f 1.. . .a f 7‘ ff) (m ‘11 U; 45 molecules based on size and configuration, but will also adsorb pref- erentially based on polarity or degree of unsaturation. In a mixture of molecules small enough ‘to enter the pores, the less volatile, more polar, or more unsaturated the molecule, the more tightly it is held within the crystal lattice. MS are available in a wide variety of types and forms. For the adsorption of CO; and water from a stream of gas, Union Carbide MS type 5-A (Hersh 1961) and type 4-A (Johansson 1963) are used, and recently a MS especially designed for this purpose was patented (Clarke, R. G. 1964). Clarke, R. G. (1964) developed a gel for the purpose of adsorbing carbon dioxide and water from environmental atmosphere that can be readily regenerated by drawing a vacuum, supplying heat or by using a combination of these two methods. This gel was made of K;CO3, Mg (N03); 2H;O and Al (NO3)3- 9H;O. A gel of MgO- Al;O3 co-precipitate with the evaluation of CO; gas. It has been noted that the gel bodies do not lose their ability to adsorb car- bon dioxide although they pick up a considerable portion of their weight in water. It has been found, to the contrary, that the gel adsorbed CO; more efficiently when they contain water than when they are dry. In designing an MS adsorption unit the following factors should be considered: 1. Adsorption equilibria of the MS. --The amount of gas or liquid adsorbed when equilibrium is established at the critical iernpe rate a der 0: life. - to the Shape 533516 I‘Orfib: 46 temperature and pressure. 2. Adsorption rate and kinetics. --The rate at which a given material will be adsorbed by MS pellets, pow- der or beads at the critical temperature and pressure. 3. Cyclic life. --During cyclic operation, (will be discussed later) damage to the adsorbent may occur and the equilibrium adsorption capacity is reduced. MS may be used in the following types of adsorption systems: 1. Static adsorption. --When formed into wafers or other shapes, MS can be used as static adsorbers in closed gas or liquid system. 2. Single-bed adsorption. --When interrupted produce flow can be tolerated, it is sometimes convenient to use a single adsorption bed. When the adsorption capacity of the bed is reached, it can be regenerated for further use. 3. Multiple-bed adsorption. --A typical dual-bed installation places one bed on adsorption, while the other is being regenerated. When the process design requires shorter time for the adsorption step, additional beds can be added to permit continuous processing of the feed. Desorption or regeneration cycles for MS Desorption, which is the most inefficient step in the cycle can be classified into four types, which may be used separately or in combination. . ‘v “, «EU; and is returr rv diction ' tional ti sure CYI llal pre sorbate picked 1 47 1. Thermalgcle. --The MS is heated to a temperature between 400 and 600°F, the bed is flushed with a dry purge gas and the MS is returned to an adsorption condition. A cooling step is then needed. 2. Pressure cycle. --Desorption can be accomplished by a re- duction in total pressure at isothermal conditions. Because addi- tional time is not required for heating and cooling the beds, the pres- sure cycle can be designed to operate rapidly. 3. Purge gas stripping cycle. --The purge gas reduces the par- tial pressure of the adsorbed component and affects desorption. Ad- sorbates pass from the adsorbent into the vapor over the bed, are picked up by the purge gas, and swept from the bed. 4. Displacement gycles. --Use an adsorbable fluid to displace all or part of the material loaded in the bed. Increased pressure adsorption on MS Gregory _e_’_t-_a1. (196 3) mentioned a number of advantages in using increased pressure adsorption (about 50 lb. /in. , gage) in the chemi- cal industry. Firstly, as the partial pressure of CO; in a given gas stream rises, the quantity of MS required falls. Secondly, since the saturation content of water in the MS remains approximately con- stant on a volume basis, the mass ratio falls. This gives a further advantage of the use of MS by reducing the sites taken up by the water molecules and also by a reduction of the temperature of adsorption, (II C 0; ing pre 16‘." 48 since relatively less water is adsorbed. (The heat of adsorption of CO; is low compared with that of water). He also found that cool- ing is of considerable importance. Reactivation at atmospheric pressure gave advantages over reactivation at higher pressures, in terms of purge losses, heater rating required, and adsorption capacity achieved. In-bed heating reduced the heat consumption compared to external he ate rs. The use of MS in CA storage The use of MS as a CO; adsorbent in CA room was first men- tioned in the literature by Johansson (1963). Johansson used Union Carbide MS type 4-A which adsorbs water, CO;, methanol, ethanol, ethylene, acetylene, propylene, n-propanal and ethylene oxide. Since water is a very strong polar compound it has a stronger af- finity for the adsorbent than do the other components of the atmos- phere. The atmosphere from the storage chamber was drawn (Fig. 2C4- 1) through the adsorption column containing the MS. After passing through the adsorption unit, the air was forced through water in a gas washing bottle in order to saturate the atmosphere before it reentered the storage chamber. The atmosphere was cir- culated by an air pump. By adjusting the amount of atmosphere pass- ing through the adsorption column, according to the rate of respira- tion, the CO; level of the atmosphere could be held at a fairly constant JL: \\\\\\\\\\\ x {§\\\\\\\®§ / oleCUlor e es 8 / ———> «— O l 13!}!P'W9H :9 —>— mun Jiv isson 196 5) 510m \J\ Sealed Storage Chamber of 1110 ('A s_\' :111t‘dlngr'uui ..__ i /l/ Jezflgcuv oi Fig ZL‘el-l. Svhvm 50 level. The MS was reactivated by heating to 500°F; during the re- activation cycle air was purged through the adsorption column in a direction opposite to that during the adsorption phase. The control of the CO; level was satisfactory. It is believed that two companies (Tectrol Division of Whirlpool Corporation and Atlantic Research Corporation) are using synthetic or natural MS in their carbon dioxide scrubbers in their CA gas generating equipment. Tectrol is using it as part of their complete external generator. Atlantic Research corporation on the other hand is using their CO; scrubber-Arcosorb with and without their catalytic burner-Arcat. The Arcosorb is designed to handle the CO; load generated by the Arcat too. Fig. 2C4-2 is a flow diagram of the Arcosorb-Arcat (Arcagen). It consists of dual tower design which operates alternatively as ad- sorption and reactivating and allows continuous fully automatic opera- tion. Tower 1 of the Arcosorb is in adsorbing position with downward air flow. The room atmosphere passes through the adsorbent and CO; is extracted. Tower 2 in reactivating position has heaters energized and reactivation air flow sweeping the desorbed CO; to the outside air. The adsorbent is thus readied for a new adsorption cycle, and after a programmed time period the 4-way valves reverse, tower 1 is put on reactivation, and tower 2 on adsorption. As water is adsorbed in this process rehumidification is provided to maintain the desired high 51 Amos numuoE acumuoaom 0.33.5083 coach—coo “5393930 nonwomom 035.34 no Swamflv Bofim ...N «Um .wrm Emhm>m Zm0ZOU zm0>xo h_._.U._<> o w>3<> All at. «$50... <<<< Zn "330... 471.11 :~’s>’ >a < < (e (\ ._< < ..< ammZmn—ZOU unmJOOU WW I: N. 2. «95:5 200m .<. U . 20:”. 023000 ZOOM {ooh So «2 zo:<>:uNO-U-‘\ ‘\\\ .. ”ZEN/”Huifme V . . . 2... . \ /. Noo moo \ \ .uooo: 3255.5 393. 853:0 mcofiEoE 57 2D. O_x_yg§n Reduction The general methods of oxygen reduction are: a. Fruit or product generated atmosphere. b. Externally generated atmosphere (1) The addition of liquid or compressed nitrogen or carbon dioxide. (2) Hydrocarbon fuel burner. (3) External gas generator. ZDl. Fruit or_product generated atmosphere After the CA room is closed, the apples continue respiring which requires 02 and simultaneously produces C02. There is a decrease in 02 level and an increase in the C02 level. As the respiration con- tinues the C02 level will reach the controlled condition (5 percent for example) and until this point is reached, in a relatively gas tight storage, the total material balance has not been disturbed volumetri- cally and the materials have only been changed in composition. How- ever, as respiration continues, the CO; level will rise above 5 percent the excess C02 will be removed by scrubbing, and the material balance on a volumetric basis is changed. Pflug and Dewey (1956) and Pflug (1960) discussed thisproblem extensively and developed a procedure for a numerical calculation of the rate of oxygen reduction. Pflug used the following variables and units: 58 Symbol Storage room volume (gross) V Quantity of fruit in storage B V Space factor Z = 1—3 Respiration rate r Oxygen level in storage Y Controlled COZ level in 0 storage Leakage or ventilation rate x Temperature T Oxygen in standard atmosphere Carbon dioxide in standard atmosphere Bushel of apples I‘ime t Oxygen diffusion ratio w Units cu. ft. bushels (bu .) cu. ft. bu. cu. ft. 0; consumed 1000 bu. day C11. ft. 02 cu. ft. atmosphere CU. ft. C02 cu. ft. atmo sphere air changes (based on empty room volume) 24 hr. °F 0.21 cu. ft. cu. ft. atmosphere 0.0003 cu. ft. cu. ft. atmosphere 45 lb. days cu. ft. Oz added cu. ft. CO;, removal It was also estimated that l bu. of apples occupies a volume of 0.74 cu. ft. (the box was included in the calculation of the volume). The oxygen level Yt of a CA storage at any time t is given by the equation Ti u“ sit, 59 59 Oz O-> t daysday Y = 0.21 (ZDl-l) where the change in oxygen level was evaluated on a day basis: AO; _ r-Oz added through leakage - 02 added through COZ absorption day — volume of atmosphere (2D1-2) The summary of the relationships of Pflug (1960) for the caustic soda ab- sorption system also will be true for any other absorption process in which there is no oxygen added by absorber through diffusion. _ _ _ cu. ft. Oxygen consumed and CO; produced — r — 1000 bu. day Oxygen added through air leakage 1000 _ _ _ _ cu. ft. ——B Vx (0.21 Y) —1000 Zx (0.21 Y) — 1000 bu. day Ox en added throu hCO remov 1— o 21 (r - 1000 z x C) — cu' ft' yg g 7- a ‘ ' ' “1000 bu. day Volume of atmosphere per 1000 bu. = 1000 (Z - 0.74) 2 cu. ft. A02 _ r -1000 Zx (0.21-Y) - 0.21 (r -1000 Z x C) (ZDl-3) day ‘ (1000 z - 0.74) A02 [ 0.79r 2x Y=O.21- =o.21- E -——-————— t day O»tdays 1000 (Z 0.74) (Z 0.74) (0. 21 - Y - 0. 21C)] (2D1-4) In the same way, Pflug (1960) shows a way to calculate the effect of oxy- gen added through diffusion, as with water scrubber. This enables him to show the effect of the water scrubber on the 0; reduction. The general equation derived: Using. reduce The ra' sorptr ITithe 1361‘, C lated 1 60 E 0.79r (l-w) Zx Y : 0,21- [ - ——-——— (O.21-Y-O.ZlC-O.79WC)J t - ‘ Z 1: days 1000 (Z 0.74) (Z 0.74) (2D1-5) Using Eq. 2D1-5 Pflug shows the following effects of: 1. Room leakage rate. The higher the leakage the longer the time to reduce the oxygen content. This shows the importance of a tight room. 2. Room fullness. The lower the space factor Z the shorter the time needed to reduce the oxygen. This shows the importance of completely filling the room with apples, the disadvantages of leaving a space between the apple boxes, and the advantage of using a pallet box compared to a bushel box. 3. Type of scrubber. Oxygen is added through diffusion, therefore, the rate of oxygen reduction is slower with the water absorber. The ab- sorption will take about 20 percent longer to develop a 3 percent oxygen atmosphere than when chemical absorption (caustic for example) is used. In the adequately tight, fully- loaded room that means 2 to 3 days longer. Table ZDl-l shows the effect of leakage, fullness and type of scrub- ber, on the time required to develop the desired atmosphere as calcu- lated by Pflug (1960): (See Table ZDl-l, p. 61). ZDZ. The addition of nitrcgen or carbon dioxide As was mentioned in the last section, the CO; which is scrubbed away is replaced with air which in turn contains 21 percent oxygen. This is not desired since a fast reduction of oxygen is needed. One way TABLE II Type O: absor; syst L \Vater raUSth SOda ébSOrp 61 TABLE 2D1-l. Time required for CA storage room to reduce oxygen level to 0.030, carbon dioxide level to 0.025, as a function of type of carbon dioxide absorption system, leakage rate .and room fullness Days to reduce oxygen level to 0.030 Type of CO; Space absorption factor Leakage rate, x, air changes per day system Z, cu. ft. ' . bu. ‘ 0 0.01 0.02‘ 0.03 2.5 13 ‘15 20 over 25 Water 3.0 19 '23 over 25 3.5 24 over 25 Caustic ‘ 2.5 ll 13 15 20 soda 3.0 16 19 25 over 25 absorption 3. 5 21 over 25 topre Itlafi thern final: OfSIC bahn finue *hrou is k6} react (Soufl [\J (It) {1 1 (J) (D O 62 to prevent this is the addition of NZ. In this way the CO; level is held relatively constant, the apples continue to bring the 0; level down and the material balance is further maintained by the addition of N2 until final process conditions are reached. When the final process conditions are achieved, after 10-20 days of storage, air must be added instead of NZ to maintain the material balance (usually the normal leakage is sufficient) since the apples con- tinue to respire and require the presence of OZ in the air. The addition of the N2 is done by releasing N2 gas from a cylinder through a pipe leading to the CA. room, or the breather bag (Sec. 4B) is kept inflated by adding nitrogen until the proper oxygen level is reached at which time the breather bag may be kept inflated with air (Southwick and Zahradnik 1958). N 2 flUShing A more recent process is the flushing of the room with N2. This can be done in either of two ways: 1. The use of gaseous NZ to flush the oxygen from the room to obtain a rapid oxygen drop. 2. The use of liquid N; as a supplement to mechanical refrigera- tion in the precooling of the apples and at the same time to flush out the oxygen from the room. The difference between these two methods is that in the first one gaseous N; from cylinders is used for flushing only, there is no cooling 63 effect, and in the second, liquid N2 from cylinder is used for pre- cooling and flushing and then gaseous N2 is used from the same cylin- der to regulate the atmosphere. There are different engineering problems with the gaseous and liquid N; involved. We will discuss only the liquid N; process and analyze the difference between the two only economically (Sec. 2J). Smock and Blanpied (1965) conducted a small scale (1500 bu.) experiment using a modified version of the "Polastream" intransit refrigeration process of the Linde Division of Union Carbide Corpora- tion (Kane 1961) (Fig. ZDZ-l). For the cooling phase of the experiment they introduced liquid N; through a ceiling mounted spray header set up in front of the blower on the evaporator. A thermostatic control in the liquid nitrogen line caused liquid nitrogen to be introduced into the room when the air temperature rose above 30°F. Since a great deal of nitrogen gas was introduced into the room during cooling, the pres- sure was released by opening the port hole in the gas tight door. After the cooling and oxygen flushing operations were finished, the unit was switched over to gaseous nitrogen introduction into the room. This meant that the liquid nitrogen had to go through a vaporizer attached to the storage tank prior to entering the room. An oxygen sensor was placed inside the CA room. When the oxygen level rose in the room above 3 percent concentration, the oxygen sensor caused a solenoid - valve to open and nitrogen gas entered the room. Since this introduction of r,_} A 2:. /. 5313mm. :::,.r_;m_._ .Z ,::r.._:..,q_c.m, ~11..- v.7; 2 5m; .L . . . mmz_<.rzoo Zmoomtz 0503 m>4<> Q_Ozm40m mqud mqo w>J<> m>4<> xooo>mk ozExammmmd J 1:3... 52533 \1 Exam zmmuomtz 9:9 1: \J\ IW/ \fl l./ \J‘ J1 mid» . / \\, _ / Cd _ . . d I. II \\ (I \ I E \ 90253 , 11 \ 4139‘ 4.1““ moaqo mmammmmd Q50: .3: c: c_\\ cf 41 4. d .v c. mmodmr >4<> 13K 930: ZOpomzzOu 0231:“. 050: 65 N2 raised the pressure in the room, the excess pressure was released through a 2-in. pipe. The end of this pipe was immersed l/Z-in. in a container of water so that not. more than 1/ Z-in. water pressure would build up inside the room. The C0,; was controlled by running pipes to and from the CA room into a dry lime scrubber. The precooling with liquid N; was very efficient. The maintenance of the oxygen level with the automatic nitrogen supply also was very efficient in a "leaky" room. (Naturally the bigger the leak the bigger is the consumption of N2.) When the room was opened and half of the apples removed the desired oxygen level was reached a day after resealing using the automatic nitrogen supply. During the period that the room was half full, twice as much nitrogen was required per day to hold the O; at 3 percent. Smock and Blanpied (1965) summarized the advantages and disadvantages of nitrogen CA rooms: Advantages: 1. Flushing out the oxygen in a CA room is very rapid when liquid nitrogen is used for cooling as well as flushing. 2. While flushing with liquid nitrogen, there is a cooling effect that can supplement refrigeration. 3. Very close, automatic control of the oxygen level was obtained by the use of the modified "Polastream" process. 4. No servicing of the equipment was required once it was in- stalled and started up. There are no moving parts. Negligible 66 electricity is required for the oxygen sensor and to activate the sole- noid valves. 5. There is no continuous flushing of the room. The apples are respiring and nitrogen is introduced into the room only to hold the oxygen at 3 percent. 6. The CA room can be opened and some of the fruit removed with quick regeneration of the atmosphere following resealing. Disadvantaggs: 1. During liquid nitrogen cooling there is a slight danger of freezing the fruit since nitrogen vaporizes at -370° F but it could be avoided. 2. When the room is not fullthe flushing rate increases and this introduces the problem of the maintenance of a high enough CO; level. 3. There is some loss of liquid nitrogen during long storage. The economical study of this system is in Sec. 2J. As the ex- periment was done for a 1500 bu. -room, definite conclusions from the results have to be judged carefully for a larger room. The addition of carbon dioxide Another possibility for a combination of oxygen pulldown and pre- cooling is the use of dry ice or liquid carbon dioxide. The rate of sublimation or heat absorption of dry ice (at the temperature of sub- limation; -110°F) is too slow. The dry’ice could be crushed to reduce 67 the particle size and expose more surface area to increase the re- frigeration capacity, but the pulverizing operation reduces the over- all efficiency therefore liquid carbon dioxide is more suitable for this purpose. There are at present "in- transit" refrigeration units similar to the "Polastream" discussed, using liquid CO; instead of liquid N;. 1 A similar modification as with the Polastream is possible. The big disadvantage is that high CO; level will be reached and extra scrubbing of CO; will be needed. However, this disadvantage is an advantage when high concentration of CO; is needed, over 20 percent as with cherries and black currants for example. This operation is not suitable for apples and will not be discussed any further. 2D3. flydrocarbon fuel burner Another way to accelerate the natural reduction of oxygen by respiration, is to burn the oxygen of the storage atmosphere with gas. Gases with relatively low concentration of N; and C0; are the most prac- tical sources for this purpose. Fig. 2D3-l shows the volumetric rela- tionship between O; and CO; that can be expected with a combustion system using air as a source of O; (Lannert 1964). Two gaseous fuels 1Liquid Carbonic-Division of General Dynamic Porta-CO; LD--and Pureco CO; Blast Chilling--Pure Carbonic Division of Air Reduction Company. 68 are shown; natural gas, which is assumed to be methane-CH4 and propane-C3H8. Fig. 2D3-1 also shows an oval which embraces the O; and CO; concentrations commonly accepted in the CA apple stor- age. Since the curves do not intersect this oval, it is obvious that CO; must be removed from the room. From Fig. 2D3-l if 3 percent O; and 5 percent CO; are desired, for example, we find that for com- bustion of the oxygen down to 3 percent the CO; generated will be over 10 percent, therefore, this excess has to be removed. The excess CO; produced is removed by an absorber as the room atmosphere is recirculated through it. By operation on a recirculatory basis the need is to convert only the 21 percent oxygen of the air in the room (after sealing the room) and the 21 percent oxygen of the air intro- duced to the room because of CO; scrubbing and leakage. When the rooms are made as tight as possible the use of the burner and in turn the absorber will be smaller which is desirable. A. full storage room (low space factor) takes less time to pull down, than does a half-full room, as more of the air space is taken up by the apples. The apples act as a supplement to the burner operation. The burner can be adjusted to accommodate for the original oxygen present in the room, oxygen introduced because of leakage and oxy- gen introduced because of CO; removal. Two companies, Tectrol Division of Whirlpool Corp. , and Atlantic Research are using catalytic burners in their CA generating equipment '69 NN ON m. “a 333 «pound 1: hm 5 sown—5800 809“ 36:66 .conhwo .m> comma .almsoaoomd zofimamzoo 2. No 6. z m. o. m m a ‘u _ o 1‘ wmmMIdm02k< AENVV mo 3: ...u coca. .— 1 umzoEnzou 1 oo 1 «2.-» 230933.. 1 do on I P 02.13: . 1 3 20:52.5: 2: 1 1 1 co. h2m5u¢3mzmt .anxLEm onczawcfizc medmOhw wmwrdwosid omqqomhzoo x w>mw4m 36:35:05 .20.. 53:35:06 coy/{C 6-20 mm. x . r All :3 mbzm> oom EoddszWX. ~. I I Ed Imwmm .Em Saaam ll .35.... 1.015.200 1.me mmzmnm \mfiim oEfimozmuE oz Wzowp “ . H253 _ I _ mas/5mg m>mmnm simlzflnwwupflzmh.’ .Iufimyebmk ‘11111111. 7 fill 2.25.. ’11 11111 1, .d 5?“.me \Iu'l.ll".l” aIIIIIIlIn‘s 2 s......1.1|.11/ \dé» .Em fu/vbhuk $23 2 v mudox . mw V £25928 025282862308 .: mxmxm mmmmommd Noo 75 5. For a recirculating atmosphere the burner is only needed for the pulldown periods. 6. When the burner is operating, some air leakage can be tolerated. Disadvantages of hydrocarbon fuel burner: 1. Excess CO; is formed and has to be removed. 2. Causes heat input into the room, therefore, more refrigera- tion is needed. 3. The constant change of fuel-oxygen ratio in the recirculating atmosphere (because of constant O; reduction) may present a problem of burner control. 4. Possible problems with toxic impurities--when the 0; level drops below the stoichiometric level, unburned fuel and impurities will be present in the combustion product. This problem is impor- tant since during the pulldown the fuel- oxygen is constantly changed because the O; level drops continuously especially with the regular burner. 2D4. Externalfis generator The oxygen reduction andthe control of the atmosphere is made by introducing externally generated atmosphere into the-room. This is discussed in Sec. 2E. 76 2E. External Gas Generator The external generation of the atmosphere (EGA) for the CA storage of apples is a different approach to atmosphere control in a CA storage room. The desired atmosphere is produced outside the room and then introduced into the room. In this way the atmosphere is not as dependent on respiration as with the regular CA. There is no oxygen reduction here because the original atmosphere actually is replaced and equilibrium is reached with the respiring apples. Tectrol Division of the Whirlpool Corporation in 1962 introduced their Tectrol (Total Environmental Control) system which is an externally generated atmosphere. Fig. 2E-1‘diagrammatically represents the atmosphere control factors affecting theistorage using their system. The storage atmosphere is a function of (1) air leak- age in (2) atmosphere supplied by the generator (3) respiration of the apples (4) atmosphere leakage. This technique utilizes the CO; pro- duced by the apples along with a constantly supplied generated at- mosphere that will complement the air infiltration to provide the desired atmosphere. When the required atmosphere is to contain for example 3 percent 0;, the generator output must be somewhat below this value so that it will be complemented by the 21 percent O; in the infiltration air. Fig. 2E-2 is a schematic of the Tectrol generator which produces .‘77 o~z+~811~o - .woaowmd ammOHm .. _ - _ _ . A wo3 and .3238 m2. Jomhowh ”6,2523 m. V 3423.. m3 Ill-I!m»\\\\\\\\\\\\\K\\\\\\\\\\\ mum—24:0 wodmodb ow._.fiz a E ..de 79 the atmosphere that is supplied to the storage chamber. Hydro- carbon gas is used in the burner for the reasons described in Sec. 2D-3. Gaseous fuel and outside air mixed together in the proper proportions are fed into the burner constantly all season. The mixture is burned catalytically with the O; in the air combining with the fuel to produce CO; and water. This step is different from the one used in the Arcat where the burner is burning the O; in the re- circulating atmosphere. Changes in the gaseous fuel input mixture result in a change in the O; content of the effluent (present an engin- eering problem for Arcat). As outside air and specific type of fuel is used (usually liquid petroleum- LP fuel as natural gas), there are only minor feed gas fluctuations and the effluent gas composition is steady. In this way there are no problems of traces of unburned or partially burned components when an atmosphere of O; lower than 2 percent is desired due to insufficient O; in feed gas. The mixture is burned catalytically with the O; in the aircom- bining with the hydrocarbon fuel to produce CO; and water. The burner effluent then proceeds to a condenser where it is cooled and some of the water is removed. The effluent then enters the scrubber section (probably MS type, see Sec. 2C-4) where most of the CO; is removed and the resultant effluent is then expelled from the genera- tor to the storage room. A valve is arranged in parallel with the scrubber to control the CO; output by control of the amount of 80 effluent bypassing the scrubber. The particular Tectrol generator Model 2197-G has a limit of 110, 000 Btu/ hr. and volumetric output of approximately 950 cu. ft. per hr. , if combustion is stoichiometric. Cooling water must be available--3. 5 gpm (70° F and 25 psig). The gaseous fuel flow will be approximately 44 cu. ft. per hr. of propane, or 110 cu. ft. per hr. of natural gas. This unit will take care of between 12, 000 and 18, 000 bu. depending on the leakage rate of the storage room. The pulldown period is about 3 days and the Tectrol generator can tolerate up to 0. 5 air changes per day. The use of the Tectrol generator has been on a lease basis only but now is offered for sale too. Advantages: 1. Fast 0; reduction--allowing reopening and closing of the CA storage room during storage season. 2. Enables operation of partially full room--part of the fruit can be removed and room resealed. and kept at the desired atmosphere. 3. It is possible to keepO; ldvel as low as 1-2 percent when desired. .- 4. Can tolerate some leakage because of slight gas pressure developed which enables possible construction cost saving and rela- tively easy conversion of existing refrigerated room to CA. 5. Automatic control of constant atmosphere. 81 6. No odor build-up because of the continuous flushing and the use of MS type scrubbers. 7. Definite prevention of impurities and unburned fuel because of the use of outside air for the burner and the continuous flushing. 8. Available also on lease basis. The atmosphere is guaran- teed and there is no operational responsibility. There is no capital investment nor expensive start up and the full cost is tax deductible rather than capital investment. Disadvantage s: 1. Excess CO; is formed and has to be removed. 2. Causes heat input into room, therefore, more refrigeration is needed. 3. The burner is working all season long and is not so flexible in use as the Arcat. 4. Does not utilize all the O; reduction provided by the apples respiration. 2F. Complete Systems for the Control of the Atmosghere Every combination of a CO; scrubber and oxygen reduction system is suitable as a complete system for CA, as shown in Table 2F-1 and in addition Arcagen, Tectrol and Eaves "Oxygen controller. " As an example of a complete CA system the Tectrol was 82 >4 ><1 N N N N NNNNNNNN {>4 K >4 N N ><1 N N mcmpnama o>$oo~om mod/3m 8350302 683 Ingm— m onwawflocmfi m mous; mcog EOmZ EOM scum? ponnspom Em 033:0 muondmoeuw mcfimasofioom goshsm mczooo + demsw ...z 3:34 33?: NZ cofimfiamom 12ng qofi osboh comme 800% <0 m 5 whonamoaum 2.3 .«o dobcoo 65 new 839$ 339880 how mcome—QEOQ mzmmmom A ...mN Mdmma. 83 described in Sec. 2E and the combination of liquid N2 and dry lime in Sec. ZDZ and I will analyze here the general overall operation of the Arcagen and Eaves " Oxygen controller. " Fig. 2C4-2 is the flow diagram of Arcagen of the Atlantic Research Corporation. The room atmosphere passes through the Arcosorb (described in Sec. 2C4) and the C02 and water are ex- tracted. The adsorbed C02 is exhausted to the atmosphere during reactivation and the dual tower design allows continuous fully automatic operation. A portion of the scrubbed gas then passes through the catalytic burner, Arcat (Sec. 2D3), where 02 is con- verted to CO; and water using liquid petroleum gas as fuel. The excess C02 produced in the oxidation process is removed by the absorber as the room atmosphere is recirculated through the equipment. The water vapor produced assists in the rehumidifi- Cation of the process stream. The catalytic burner-is bypassed When 02 pulldown is not required; the regulation of the atmosphere is made easily by automatic operation of the scrubbing and the regulation of the percent of atmosphere going through the burner. Fig. 2D3-3 is the diagram of "Oxygen controller" for CA, as was suggested by Eaves and Lockhart (1964). The apparatus essentially consists of a propane burner (Sec. 2D3) cooler con- denser coil, and an independent dry lime scrubber for the removal Of the CO; formed. Toxic carbon monoxide was formed when using 84 the burner on the closed recirculating atmosphere during the pre- liminary tests. By disconnecting the return pipe to the burner this hazard was eliminated. In the future the air discharge pipe will be put through an outside wall thus eliminating any possible hazard from carbon monoxide. By circulating the atmosphere from the burner through air cooled coils, heat is removed. The cooling coils also remove the moisture produced by the propane combustion. The excess CO; which is formed by the burner and apple respiration is then absorbed in the dry lime absorber. 2G. C02 and 02 Measurement and Control §ystems To control the composition of the atmosphere in the room we have to measure the carbon dioxide and oxygen concentration. There are three general measurement techniques used in the CA analysis: 1. Chemical methods. The simplest methods are chemical. CEll‘bon dioxide and oxygen are successively adsorbed from the gas Sarhple, and observations are made of the change of volume, or pressure. Example of the constant pressure and temperature and Changing volume system is the Orsat apparatus. The Orsat type C02 and Oz measuring apparatus is relatively inexpensive, and is of such Construction and operation that it can be operated successively by S130rage operators of all training levels. It is the most common atlalyzer used today in CA storages. Smock and Blanpied (1960) 85 described the apparatus and its use in CA storages in detail. The constant-volume, changing pressure types are inherently more accu- rate but are more expensive in use. The chemical methods are time consuming and do not lend themselves to easy mechanization. 2. Measurement methods based on some physical properties of the component to be analyzed. These methods are much faster than the chemical ones and easily mechanized for automatic operation. Physical properties such as thermal conductivity and paramagnetic susceptibility are the most common properties used. Some other properties such as absorption of monochromatic radiation, speed of Sound in the gas, changes of density and infrared have been used too (Fidler 1961b). Thermal conductivity Detecting changes of thermal conductivity (using either hot wires, 01‘ thermistors), is a potentially accurate method to measure C02 COncentration. With change of carbon dioxide concentration, the thermal conductivity is changed and the instrument can be calibrated to read COZ. As an example Beckman Model 7C thermal conductivity tyDe carbon dioxide analyzer will be described as used to measure the C02 in the CA room (Hunter 1965). This analyzer measures the con- Clentration of one gas component in a mixture of gases, utilizing Specific sensitive thermal conductivity filaments as detector for CO; in a mixture of C02, 02 and N2 in our case. Since the thermal 86 conductivity of C02 is much greater than that of N2 or 02, the pres- ence of a small amount of C02 will be easily detected. One advan- tage of using this analyzer is that it can detect ammonia gas leakage from therefrigeration system- It happens that ammonia gas has also a high conductivity and the apparatus will act both as C02 analyzer and as an ammonia leak detector alarm system. In normal operation, the CO; concentration will not change rapidly, but the development of a small ammonia leak will cause the C02 reading to increase rapidly which will indicate that something is wrong in the room (Hunter 1965). The thermal conductivity analyzer can be used to measure oxy- gen concentration also, with the addition of change over valves and a Small carbon furnace. The procedure is first to observe the per- clentage CO; and then to operate a change-over valve, passing the Sample through a carbon furnace. In the carbon furnace the 02 in the sample is burned to C02, the sample then passing through the thermal conductivity analyzer again and a new reading is obtained. The 02 concentration is easily calculated (Mann 1960). Paramagnetic susceptibility Oxygen is unique among gases in being strongly paramagnetic. Thus, by measuring the. magnetic susceptibility of the CA gas its 02 content can be determined. No other gas in the CA atmosphere can affect the reading significantly. When the gas sample containing QXYgen is drawn into the test chamber, the magnetic force is altered, 87 this change is proportional to the oxygen concentration in the sample when the proper equipment is available. The measurement is simple, accurate and fast. The difference among paramagnetic oxygen analyzers is in the sampling method (automatic or manual) and the possibilities to record the results. As an example the Beckman D2 indicating oxygen analyzer for 0-25 percent 02 range ($295) for manual sampling, the C2, E2 can be arranged for continuous oxygen analysis of sample stream but not for recording (range $660- 1200) and F3 is available optionally with complete sampling systems, auto- matic control and recorders (the price of the complete system is over 55 2, 000). Hunter (1965) reported the successful use of Model F3 with Sampling and recorder for 325, 000 bu. apple storage. 3. Polarography (e_1ectrochemica1 method). An oxygen sensor Was developed, which is basically a polaragraphic oxygen electrode. In this method, a measured small potential is impressed across a pair of electrodes one of which is a non-polarizable reference elec— tr‘Ocie, the other a polarizable inert electrode. The magnitude of the Current which flows is proportional to the amount of oxygen-present in the sample. The Beckman model 777 and 778 process oxygen analyzers can be used to measure the oxygen in a CA room. The sensor's small Size makes it ideal for installation in gland assemblies and thus 88 suitable also for control systems. A sealed gland assembly is available which permits the sensor to be inserted into each room if needed. The current produced is read out or can be arranged as output for potentiometric recorder or control system. The readout may be directly calibrated in oxygen partial pressure, or in percent oxygen. The control of the gs level in the atmosphere The control can be done in four ways: 1. Manual operation. Carbon dioxide and oxygen analysis is followed by manual operation of C02 scrubbers, air fan, 02 or N; addition. In this case the C02 and Oz analysis can be done by any method mentioned and the control is manually on a trial and error basis. 2. Prmfl'ammed operation. Intermittent electric time clocks are used to program the operation of the scrubber, air fan, 02 or N; addition. The CO; and 0; analysis can be done in any method men- tioried and the program is adjusted according to the gas analysis. The program will be different for the oxygen reduction period and for the storage after equilibrium of the atmosphere is reached. The adjustment is simple; if for example under any given schedule, the C02 tends to increase slightly from one day to the next, an increase of the operating time is needed and vice versa. The adjustment 89 of the operation of the air fan, 02 or N; addition is made in the same way. Most of the scrubbers are operating on a programmed opera- tion, today. 3. Control of the atmospheric generator. The control of the atmosphere generated by the Tectrol generator is made by adjusting the generator. The adjustment is based on the 02 and C02 analysis of the storage room atmosphere. 4. Instrumental control of the CO; and Oz. Zahradnik (1964) mentioned the idea of complete instrumental system of indication and control of the C02 and 02 levels in the CA room operation. The C02 indicator-controller controls the operation of the scrubber as re- quired and the 02 indicator controller controls the addition of air, nitrogen or oxygen to the room. Control of the carbon dioxide level. The control of the con- stant C02 level as measured by thermal conductivity analyzer can be done in few ways namely; (a) using a mirror galvanometer and a Photocen acting on relays. --When the cell is hit by the light spot, the relay opens a solenoid-valve which operates the C02 scrubbing unit. The scrubber operates until the CO; concentration drops just a bit below the value to be kept constant (Wolf and Kuprianoff 1961), and (b) when recorder is used--the recorder is provided with a retransmitting slide wire which can be used for control purposes. Control of the oxygen level. The oxygen concentration is 90 measured by the paramagnetic oxygen meter or the polargraph 0; sensor and small relay and solenoid valve feeds the system with air, 02 or N; or operate an 0; burner, as desired. When the oxygen content becomes just a little lower than the desired value, an elec- trical impulse is givento open the solenoid valve, this valve is closed after the 02 content has raised sufficiently. The oxygen sensor is used in the modification of "Polastream" (Smock and Blanpied 1 965) to introduce N2 gas into the room when the 02 concentration is too high. Fig. ZG-l shows schematically the possible C02 and 0; measurement and the control of liquid nitrogen system. Wolf and Kuprianoff (1961) developed an apparatus for the control of the storage and simultaneous measurement of the gas exchange of the stored This apparatus is a laboratory scale one, using independent produce. automatic regulation of the concentration of the 02 and C02 in the internal gas atmosphere. The CO; concentration is measured by thermal conductivity analyzer which also controls the C02 level by absorption in a solution of potassium hydroxide; 02 concentration is measured by a paramagnetic oxygen meter, which by means of small relays and solenoid valve feeds the system with oxygen flowing from a Special gas container. figussion of CO; and 0; measurement and control The problem of C02 and 0; measurement is whether to use 01‘ sat or a combination of C02 and OZ instrumental analyzers and 91 recorders system, and also, which size storage (how many rooms or measurements) will justify the use of expensive instrumentation. Fig. ZG-l is a schematic instrumentation system for multiple CA rooms. Any multiple system will require one 02 analyzer, one CD; analyzer, at least one recorder and a sampling device for con- tinuous atmosphere analysis. In addition a switching system is needed to operate the system and its price will depend on number of rooms. An approximate analysis based on comparison of annual cost of operation of Orsat and the above system is given in Appendix 2- 1. The Orsat is time consuming measurement and the Orsat service life is relatively short (assumed to be 3 years) on the other hand the instrumentation system is expensive but is automatic. The instru- mentation system price was approximated for a relatively simple System based on "local" design and also on a specific proposal from Commercial contractor (Beckman Instruments, Inc. ) for a complete System. It was found that with more than 4 rooms with the "local" System and more than 7 rooms with the commercial system the in- StI‘umentation system is more economical than using Orsat. This calculation is based on the fact that we pay for the time the Orsat measurement is made as regular work. Some operators, esDecially small ones do not consider their time as money and believe that an operator must be present anyhow. In this case the N .2553; ma: mum own“: ham Seaman pod mpcmoc 1.50th 63x30 conpmo .30 $00 Hmscc< .Thm Hamish. 115 no.“ “moo H9535. ..Hquwcocow HofiooB 45 000 .0N one mo ems 65 co comma. mm? .59. 000 .2 .000mw on 8 0955mm was? .56. 000 .0N no.“ hopmgocom 3.5068 00 $00 33qu c 33:55. «soar. .Sn 000 .00 9: mo ems 05 so comma. mag 45. 000..0N can 000 .3 so.“ «moo Hmscdiaqo -Lmoo mcfioooopo so w5>wm manammom mo omsmomn— poemsflpm mm? emoo Hagan chmsfl NZ 3502 one. .3 -N 50.2953 wnfiooo cam mGEmsfl mo cofimgnsoo magma o. .CBOEHSQ 25 mo cmoo 65 ohm mwmofiaopma E mnoQESZd $92 2. 25.88.18.833 3.0.33.3. Stews afocmma gems 80.3 so .2 E. :58 «m; at .3 om .N as .3 mmh a .3 we a: .o; -.N $.00 ~.m 08.8. a: .3 0m. 25.88.18.238 30.: mm .m a. .3 m .3. a: 0-0 mm .N 3.8 N .m 08.2 mam mamaoca .0.chqu 00 .N .5833me has \Eoo H .cofimhoao omawq 30.3%. 30.33.813.88; 0.0004. acme 3.7;; 3.89.” 35.8 893 mm. 30.: vim 353 film $.00 04. 3.: we 8.70 N; 3.8 ~.m ooodm $933.. Sedatiofiwite $.00 04. acme 3.3-3.3 Sena 80.2 mmm ocmaoca A: $280 w .0 $0,038,620 :22 E356 m .qoflwcoao imam NZ 3:33: 833:8 3.. Nz .0: cc .8 8:356 9% 95:25.“ mnwgmdfl abacus. .8925 .5925 «6on:3qu mqfimsfi vehicle. mnmnmsfl meadow w mm @3300 um ~Z wcfiooo “0 NZ mo 45 mo>mm €8.34 20.589 .8380 3:34 0:33: 385 8838,. nmz 3:05 QNZ 3:34 mxmcozod 55031.5 v hem Hmnmsn you 3:03 sofiuoscoh Combs so.“ “woo Husscxw .Nuwm madmjwh. 116 TABLE 2J- 3. Annual cost in cents per bushel for combinations of carbon dioxide removal and oxygen reduction Small scale operation 3 cents/kWh, 8. 4 cents/lb. propane gas, 37. 5 and 54 cents/ 100 cu. ft. liquid N2 . . . . . 3 Volume 02 reduction Natural L1qu1d Liquisz . N2 flushmg bu. of resp1r- . . apples CO removal ' ation flushmg adjusted 2 (54 cents) (54 cents) One pulldown 15,000 Water 1. 59 2.79 1.99 20,000 1.59 2.79 1.99 60,000 1.59 2.79 1.99 15, 000 NaOH 7.89 9.09 8.29 20,000 7.61 8.81 8.01 60,000 7.62 8.82 8.02 15,000 K2C03b >6.45 >7. 65 >6.85 20,000 >4.85 >6.05 >5.25 60,000 >4.85 >6.05 >5.75 15, 000 Arcosorb ---- ---- ---- 20, 000 (MS) ---- ---- ---- 60,000 3.52 4.72 3.92 15, 000 MEA 4. 65 5. 85 5. 05 20,000 4.65 5.85 5.05 60,000 4.65 5.85 5.05 15, 000 Dry-lime 1.47 2. 67 1. 87 20, 000 based on 1. 47 2.67 1. 87 60, 000 3 cents/bu. 1. 47 2. 67 1. 87 const cost 15, 000 Dry-lime 1.29 2.49 1. 69 20, 000 based on 1.29 2.49 1.69 60, 000 1. 5 cents/bu. 1. 29 2. 49 1. 69 const cost aThe liquid N; flushing annual cost was adjusted because of possible saving on precooling costs using combination of flushing and cooling (Appendix 2- 3). b K2C03 scrubbing cost is based on initial price--suggested re- tail price in Europe without shipment and installation (Appendix 2~2). 117 TABLE ZJ- 3 (continued) Liquid Liquid N,a N2. flushing Arca t Eaves Tectrol flushing adjusted burner burner as a (37.5 (37.5 burner cents) cents) One pulldown 4 pulldowns 2.39 0.39 5.76 4.03 7.27 2.39 0.39 4.83 4.03 6.11 2.39 0.39 2.94 4.03 6.11 8.69 6.69 12.06 10.33 13.57 8.41 6.41 10.85 10.05 12.13 8.42 6.42 8.97 10.06 12.14 7.25 >5.25 >10.62 >8.89 >12 13 5.65 >3.65 >8.09 >7.29 >9.37 5.65 >3.65 >6.20 >7.29 >9.37 4. 32 2. 32 See Arcagen 5. 96 8.04 Table 2J-5 5.45 3.45 8.82 7.09 10.33 5.45 3.45 7.89 7.09 9.14 5.45 3.45 6.00 7.09 9.60 2.27 0.27 5.64 See Eaves 7.15 2. 27 0. 27 4. 71 0; controller 5. 99 2.27 0.27 2.82 Table 2J-5 5.99 2.09 0.09 5.46 See Eaves 6.97 2.09 0.09 4.53 0; controller 5.81 2.09 0.09 2.64 Table 2J-5 5.81 CInitial cost of Tectrol generator for 20, 000 bu. was assumed to be $5000. 118 TABLE 2.1-4. Annual cost in cents per bushel for combinations of carbon dioxide removal and oxygen reduction Large sgarle operartion 1 cent/ kwh, 2.83 cents/ lb. propane gas, 37.5 cents and 54 cents per 100 -cu. ft. liquid N2 , . o o c a Volume 02 reduction Natural qumd qumd.Nz . N2 flush1ng bu. of resp1r- . . apples CO removal . ation flush1ng adjusted. 7‘ 454 cents) (54 cents) Onejulldown 15,000 \Nater 1.23 2.43 2.85 20,000 1.23 2.43 2.85 60,000 1.23 2.43 2.85 15,000 NaOH 7.63 8.83 9.25 20,000 7. 37 8.57 8.99 60,000 7. 35 8.55 8.97 15, 000 KZCO3b >5. 75 >6. 95 >7. 37 20, 000 >4. 32 >5. 52 >5. 94 60, 000 >4. 32 >5. 52 >5. 94 15, 000 Arcosorb ---- ---- ---- 20,000 (MS) ---- ---- ---- 60,000 2.78 3.98 4.40 15,000 MEA 2.49 3.69 4.11 20,000 2.49 3.69 4.11 60,000 2.49 3.69 4.11 15,000 Dry-lime 1.47 2.67 3.09 20, 000 Based on 1. 47 2. 67 3. O9 60, 000 3 cents/bu. 1. 47 2. 67 3. 09 const cost 15, 000 Dry-lime 1.29 2. 49 2.91 20, 000 Based on 1. 29 2. 49 2.91 60, 000 1. 5 cents/bu. 1. 29 2. 49 2. 91 const cost aThe liquid N; flushing annual cost was adjusted because of possible saving on precooling cost using combination of flushing and cooling (Appendix 2-3). b KZCO3 scrubbing cost is based on initial price--suggested retail price in Europe without shipment and installation (Appendix 2-2). 119 TABLE 2J-4 (continued) a Liquid Liquid N2 N2 flushing Arcat Eaves T:6.55 >5.57 >9.57 >7.09 >11.11 >5.12 >4.14 >7.20 >5.64 >8.49 >5.12 >4.14 >5.37 >5.66 >8.51 3. 58 2.60 See Arcagen 4. 12 6. 67 Table ZJ-5 3.29 2.31 6.31 3.83 7.85 3.29 2.31 5.39 3.83 6.68 3.29 2.31 3.54 3.83 6.68 2.27 1.29 5.29 See Eaves 7.17 2. 27 1. 29 4. 37 Oz controller 5. 66 2.27 1.29 2.52 Table 2J-5 5.66 2.09 1.11 5.11 See Eaves 6.65 2. 09 1.11 4.19 0; controller 5. 48 2.09 1.11 2.34 Table 2J-5 5.48 cInitial cost of Tectrol generator for 20, 000 bu. was assumed to be $ 5000 . 120 05005 <0 Ems mam 53505 00. 00.30500 50.5008 950: 0053 mmafigm mo mamoo 0050950000 :0 450500 ow m0 @335 :0 comma. $000 $523 50.3.0080 .0500 2000253 50 cowmwfim 20.5008 .3 00690.3 0.236: Hmhmcmm mo 025820.25 no 0025 2.000 0038.300 28 .Hofiwnmcmw .55 ooo on .305 no.“ ooomm .20 03.3 m 00 6025 2.000 .33me 3.0850538 20.50080 .5000 00305000 000 0003 no.“ 0036500500 20002553 mo 0233a 50.5008 055 m0. 00090.3 0.230: Hmhocow mo cofimfihofl: 00 00000. 3000 .Hmfiswoh A0000: 20.3.0080. 00530.» 30 50m «000 :05 .55 ooo .oo 00 00000. 050 00.000 commohdflm 8 .o om .2 4.2 o .3. mm .2. m mm.m 8.5 4.2 6.8 Sim 2 23.3 538.68 2. .N . . . . . 66.... 66865 S o 2 2 m 2 o 3 wow 2 m .nSflcmo mm .m 8m mm .m 2 .o v .2 o .2. 6mm .2 2 ooo .om 53.856 mo .2. . . . . . 2% 288.3 mo 6 ow 2 m 2 s S mom 2 m 538:8 v .m norm mm .m mm .2 m .2 s .2. «8 .2 2 ooo .2 95005 .55 o>mm hmfizmom 0>mm Mafiamm £5 00300 oooo 00 00253 p o p 5 now $500 26 .3 .HGSOHHGOU no mmw OHSQ Oh om mmwm Oh om INOHAN 5.59.5“ $85 0 638m A n I H .2. A 3 2 a. .2. -85. 2 > 053050 ohonamofium 00:05:00 8.0.3500 wcfimwxo 35950 no.“ 505025 .80 0500 5 0.000 505954 .muom m4m<9 121 000000 :0 80.02 0000 00000000, 00.0 00300 00 0.0000 .0w00000 .00 ooo .oH .00 003000300 00 00000 000 00000 000.0 0.00 odm N.mm 0.mm FAN w.om 0000 om 0000000 0000 000 0000000 .00>00000 $00.0 300.000 .0000 o0 000200 000000 00000 000.3 N.m0 N.w0 0.00 0.~N 0.0: 2.0m NZ 00:00 NZ 3:00 NZ 3:00 Nz 00:00 ...z 3:00 Nz 00:00 .0 :0 .0 :0 .0 :0 .0 :0 .0 .:o .0 :0 000 \00000 m .mm 000 \00000 «m 000 \00000 m .NM 000 \00000 0m 000 \00000 m .NM 000 \00000 0m w00>00 0000 m03000000 .00 0000000 0000030 00000 w00000000 000 w000000m 00.000 w00000000 000 w0£000000 0001000 00000000 00500.000 000 w000000m 3003 000000 00 00000 000000 00003 000 00.0 0000000 0000000050 0020,5000 00 w00000000 000: 000 000 00300030000 ..00000000000: 0 00.0 000000 .000 00.000 00 00000 000000 00000030m 0.00 "040.33. 122 Discussion of the Results of the Cost Study CO; Removal fligh utilities cost, local labor. The most economical carbon dioxide scrubbing system where utility costs are high and local labor is available are the dry lime and water systems. The Arcosorb (MS) is more expensive but much lower than MEA, KZCO3 and NaOH, at present the Arcosorb units are built for 60, 000 bu. Low utilities cost, expensive labor. The water system is the most economical; however, the difference between the cost of the water system and of the dry lime system is very small. MEA and Arcosorb are about twice as expensive as the water and dry lime system but are more economical than the KZCO3 and NaOH. Oxygen Reduction The development of the low oxygen atmosphere using the natural respiration of the fruit entails no out of pocket costs. How- ever, the time delay in developing the atmosphere can cause a loss - in quality of the fruit andtherefore has, in reality, a negative cost or liability. There has been no attempt in this study to assign a dollar value to the loss in quality that may occur during the development of the low oxygen atmosphere using the natural respiration processes of the fruit in the storage. 123 Oxygen reduction costs with Iggh cost utilities (liquid nitrogen, 54 cents per 100 cu. ft. , propane 8. 4 cents per 1b.). The discussion will be divided into two parts: 1. One pulldown only--when it is not necessary to re-open the room during the storage period. 2. Four pulldowns--allowing three additional re-opening of the rooms during the season. 1. A combination of liquid nitrogen flushing and cooling is the most economical method. The liquid nitrogen flushing, the Eaves burner and the Arcat (at its optimum capacity--60, 000 bu.) comprise the next most economical group. Introduction of a smaller Arcat burner would probably change the cost figures for 15 and 20 thousand bu. The Arcat is more economical than the Tectrol generator when the Tectrol is used only as a burner. 2. The Arcat when used at its optimum capacity is the most eco- nomical method. The Eaves burner is much more economical than the Arcat (for 15 and 20 thousand bu. ), liquid nitrogen, and the Tectrol generator when it is used as a burner. Oxygen reduction costs with low utilities costs (liquid nitrogen 37.5 cents per 100 cu. ft. , propane 2.83 cents per lb. ). 1. One pulldown only--a combination of liquid nitrogen flushing and cooling is the most economical method. Liquid nitrogen flushing, Eaves burner and Arcat (when used for a 60, 000 bu. 12.4 operation) comprise the next group. The Arcat for 15, 000 and 20, 000 bu. is more economical than the Tectrol generator for this purpose. 2. Four pulldowns--the Arcat (for 60, 000 bu. ) and the Eaves burner are the most economical, followed by liquid nitrogen flushing and cooling. Arcat (for 15, 000 and 20, 000 bu.) and liquid nitrogen flushing are quite comparable and are more economical than the Tectrol generator when it is used as a burner. Combinations of Scrubbing and Oxygen Reduction Systems The results are summarized for high utilities cost (small scale operation) in Table ZJ- 3 and for low utilities costs (large scale oper- ation) in Table ZJ-4. Small and large scale operations The costs per bushel on the small scale operation are higher than that of the large scale operation except when using a combina- tion of flushing and cooling with liquid nitrogen. In this case the saving on electricity in the small scale operation (because of the more expensive electricity for mechanical refrigeration), is higher than in the large scale operation using the same liquid nitrogen costs. The economical order of the different combinations within each group is 12.5 nearly the same. When the room does not have to be opened during the storage period, combinations of liquid nitrogen for flushing and cooling or flushing only, with dry lime or a water scrubber are the most economical. When re-Opening is needed a combination of the Eaves burner with water (or " Oxygen controller") is more economical than the Arcat with water and dry lime scrubbing. The Tectrol generator used as a burner is less economical than the Arcat. With the 60, 000 bu. the only difference is that the Arcat com- binations are more economical than Eaves' burner. Tectrol gener- ator when used as a burner is more expensive than the other two methods. Several Existing Complete CA Systems The results are summarized in Table 2J-5. Arcagen for 60, 000 bu. and Eaves " Oxygen controller" are much more economical than the Tectrol (lease) system, with and without saving on construc- tion costs. (Externally generated atmosphere systems can tolerate higher gas leakage in the CA room than a regular CA system. Prac- tically this is an advantage when a regular cold storage is converted to CA room or when a new CA room is too leaky). The introduction of the Tectrol generator on a purchase basis-- $ 5000 for the 20, 000 bu. capacity generator (Pflug 1965), will bring Tectrol to a more com- petitive price position. Assuming a saving on construction costs the 126 Tectrol system costs are close to Eaves' and Arcagen for 60, 000 bu. Even without the construction costs saving, Tectrol (purchase) costs for 15, 000 and 20, 000 bu. are more economical than the Arcagen. However, as the costs for the use of Arcagen for 15, 000 and 20, 000 bu. were based on the Arcagen initial cost for the 60, 000 bu. unit, the costs are relatively high. The Atlantic Research Corp. is plan- ning to come out with a smaller unit for 20, 000 bu. which might be more economical for this volume. The calculations for the " Oxygen controller" are based on 6, 000 bu. room but the figures are so low that a further look into this process is highly recommended. Combination of "Polastream" Modification and Dgy Lime Scrubber This combination is used as a complete system of CA for the whole storage period and its annual cost is summarized in Table 2J-6. The combination of flushing and maintenance of the atmosphere all season long is not recommended especially when re-opening of rooms during season and operation with rooms which are not completely full. The use of liquid nitrogen for flushing and cooling only is v_e§y promising but is only in the experimental stage. SECTION 3 GENERAL CONSIDERATIONS IN THE DESIGN AND SELECTION OF VAPOR BARRIER 'AND GAS SEAL MATERIALS 3A . Introduction 3A1. General There is a large number of construction materials available for use in the specific areas of controlled atmosphere storage con- struction. Since there are a number of structural or outer surface materials plus insulation materials, plus gas seal and vapor barrier materials, and since the respective materials may be combined in many different ways, there is a large number of individual specific designs of floors, walls and roofs. The most common type of con- struction scheme in use is that using concrete block or tilt up concrete with board form insulation described by Pflug§_t_a_l. (1957b), Pflug and Dewey (1959), plywood described by Zahradnik and Southwick (1958), and buildings constructed using prefabricated insulated panels by Poms (1965) and Dow Chemical (1965). In general, all of these designs have included some type of vapor barrier and a satisfactory gas seal material. In this discussion, the problem and possible solu- tions to the problem of water vapor transfer and CA gas tightness 127 128 will be considered as these two gas barriers are influenced by weather conditions, material used and different design procedures. Controlled atmosphere storages require (1) gas tight wall construction so that the design oxygen-carbon dioxide levels can be maintained in the storage, and-(2) prevention and elimination of moisture accumulation and condensation within the insulation or within the storage wall. 3A2. Water vapor barriers Water'vapor barriers are those materials which are recog- nized as retarding the transmission of water vapor with reasonable effectiveness under specified conditions. Water vapor barriers may be applied to the insulation to prevent accumulation of water within the insulation or they may be applied to the outside or inside wall to prevent moisture from migrating into the insulation. The objectives of the vapor barrier are (1) to keep the insulation dry and reduce the heat load requirements for the cooling system, (2) prevent structural damage that may occur due to rot, corrosion, or the expansion effect of freezing water. The effectiveness of any vapor barrier system in preventing water accumulation in the walls and insulation depends upon its natural vapor barrier properties, the care and thoroughness in which it is installed and also its location within the insulated section. Vapor 129 barriers must be applied to the surface exposed to the higher water vapor pressure. For most applications, this will be the warm side. The vapor barrier problem is much more complicated when a reversal of the warm side takes place due to seasonal changes which is what exists in the CA storage wall. Work carried out by Lorenzen (1963) in New York State where the operating temperature of the storages was 32° F found that the warm side of the wall is on the outside for 8-9 months of the year and is at the inside for the remainder of the year. This problem is discussed in Sec. 3C. At this point it is perhaps well to recognize that there is no material installed by man under field conditions that can be considered a perfect vapor barrier. 3A3. Thermal insulation Thermal insulation is used in cold storage walls to minimize the cost of refrigeration equipment necessary to remove the heat that flows through the walls and in addition thermal insulation is required to prevent surface condensation and therefore is a factor in the room humidity. Moisture must be kept out of insulation if the insulation is to remain effective. Any moisture that moves into the insulation area by diffusing or passing through the vapor barrier system should be able to pass through the system into the cold storage space. Some types of insulation are manufactured using a close cell system that will perform adequately without effective vapor barrier. 130 Although these materials can be used without a vapor barrier, a good vapor barrier is recommended for all refrigerated storage construction. 3B. Moisture Migration Throggh Materials The basic migration of moisture through materials of the type used for building construction and insulation takes place by three mechanisms (Lorenzen 1963). (1) total volume transfer, accounts for rain, seepage, or infiltration of moisture bearing air, (2) hygro- scopic action, accounts for the movement of moisture in intimate contact with solids, (3) diffusion of moisture as a vapor which is the mechanism of major significance and is due to vapor pressure differences. Water vapor transmission through materials under the influence of a water vapor pressure gradient is called vapor diffusion. The designed equation commonly used to describe this is: w = pAt‘A‘i‘E (313-1) where: W total weight of vapor transmitted, grains l4 permeability, grain in. /(sq. ft. )(hr. )(in. Hg vapor pressure difference) A area of cross section of flow path, sq. ft. t time of transition, hrs. 131 AP vapor pressure difference, in. Hg 1 length of flow path, in. Or using the permeance coefficient M where M = if perm or grains/ (sq. ft. )(hr. )(in. Hg), Eq. 3B-l will take the following form: W = MAtAP (3B-Z) Fortunately, in many cases consideration of water vapor pressure differences alone can be used to obtain a fairly accurate estimation of the water vapor flow. The above equations can be used to explain the design possi- bilities where reversible vapor flow, flow paths which terminate within the material, and variable vapor pressures exist. When the vapor flow path terminates within the wall, a wetting potential exists (AP is positive). This situation exists when the storage temperature is below the dew point temperature of the ex- terior air. The vapor flow path may originate within the wall and terminate outside the wall under which conditions a drying potential exists and the pressure difference is negative. This happens when the storage temperature is above the exterior dew point temperature. When the vaporflow path terminates within the wall an accumulation of moisture will take place. It is possible to store a limited amount of moisture in hygroscopic materials, the amount depending upon the equilibrium moisture content of the materials. Moisture storage capacity for a given amount of hygroscopic material is given by the equation: 132 Ws = [H1(m'1' - m'l) + H2(m'z' - m'z) + ' - -]7000 (3B-3) where W moisture storage. capacity, grains H dry weight of hygroscopic material, lbs. initial equilibrium moisture content of material, 1b. / lb. m" allowable equilibrium moisture content of materials, lb. / lb. 3C. The Effect of Weather and Reversible Vapor Flow It is necessary that both the annual and diurnal climatic varia- tions be considered in order to design a CA enclosure using gas and vapor barrier materials properly located for a specific geographical area. Lorenzen (1963) in his extensive study provides the general environmental background on which a CA structure vapor balance system can be based. Lorenzen presents the average climatic con- ditions-for the Rochester, New York area in the time series form. Lorenzen (Fig. 3C1) indicates the manner in which this time series climatic data can be used to define wetting and drying potential for the wall for a specific operating temperature of 32° F. If the potential outward movement of moisture is greater than the potential inward movement of moisture, then the wall will be comparatively dry. If the potential inward movement of water is greater than the potential outward movement of water, then the wall will become water logged. 5 ion”; :..x:.Z:._. ...».X 57:15:75: 3:323:23; _...._.::;:.:.L ..u._.::m 6.; .2: $1 .2713: ....n.:._>..:c :2... 3:29,, A}: 5.292 >42 1.32 23. >02 dem 52. _ _ _ a a . a _ _ d o Aommu no: 93980 .I R} JI_. N1 W . if” F NI / vj ' ll'lll'llll E 3:. a; ...N. a a _ <.. <.. _ A _ V _ w .—. — 41M. HO ‘ _ w v l 3 LI . 8 Lu\ mczwmmca coao> v S V .4 EmfiEd n no 3 I In. I.| N J. N W . m H n b w . u gum. H ad- “Hm—I Bow moo *0 fl mesmmmcd 53> $28.38 W H. F U T W Mi. . d r L n' "|u Aoow... “.9: oczocmaocozk _ b h _ — — P _ — _ m. 134 Diurnal climatic variations (Fig. 3C-2) are significant too and have to be taken into account. Vapor flows from a point of greater vapor pressure to a point of lesser vapor pressure. In Eq. 3B- 1, Ap =p1- p2 and p1 - p2 may be lesser pressures, resulting in i Ap, indicating that vapor flow may be in a positive or negative direction. In a structure Ap can be positive or negative, depending on the relative vapor pressures on the end of the external and internal environments. There is a reversal sign for Ap as effected by diurnal climate varia- tion, by average climatic conditions and by interior environmental design conditions. Lorenzen (196 3) analyzed this problem extensively, basing his examples on New York weather conditions and arrived at a "balancing technique for moisture control and structural components of CA storages" which will be discussed in Sec. 3E. 3D. Gas Seali_ng_t_he CA Room Smock (1962) has outlined specific recommendations for the location of the gas seal, the materials used in the gas seal, and the methods of applying the gas seal. Pflug and Dewey (1959) have worked extensively with gas sealing of CA storage rooms and have also made recommendations regarding the location and installation of gas seals. The purpose of the following discussion is to sum- marize the general principles involved in the gas sealing of controlled atmosphere storage rooms and to point to theoretical possibilities that can be adapted for use. Sci 9.2.2954, moiiméz. near...) oursria 71.35.“: :_:._m3.n._: 3!: Inzoz no >2 .Emmcoom L .09 :54 1 A W w . ' j I ' jlfi O m j. . 3: no 'N. d H _ 3398,23. B . 05.930 mm no 8 \ ad... m> Ema. 260 Wu R X . ... - M j a; l H u: 1 ...D / :4 EmBEd Co I . m.> .Eod zoo ..um 136 The gas barrier is critical for the successful operation of the CA storage. It must be relatively impermeable to oxygen, nitrogen, and carbon dioxide. Pflug (1960) calculated the rate of leakage of CA storages that can be tolerated and still permit development of the desired atmosphere, as a function of the fullness of the room and the rate of respiration of the fruit. The maximum leakage allowable for a fruit generated CA room may be as low as 0.028 air changes per day. The gas seal must maintain this low leakage rate against con- tinuous pressure differentials that may be as high as 0. 1 in. of water gage (Zahradnik e_t_i1. 1958). When externally generated atmospheres are being used, there is only an economic limit to the rate at which leakage of the atmosphere through the room can be permitted. Up to this point, we have been treating the gas barrier separately from the water vapor barrier in the CA storage. We must now recognize that the CA gas barrier is in the final analysis a water vapor barrier and is in reality a nearly perfect water vapor barrier; in fact, it is a much more perfect vapor barrier than anyone can afford to put in a storage for a vapor barrier purpose alone. Theoretically, the best place for placing a vapor barrier is at the warm side of the wall which is the outside for 32° F refrigerated storages. The gas barrier basically must be placed on the inside Wall of the CA storage which is the cold wall. Simply, the reasons for placing the gas barrier at the inside wall are that the gas barrier 137 must be near perfect, it must be accessible to repair from year to year because if the gas barrier is not adequate, then all is lost. Essentially we are saying that we cannot tolerate a poor gas barrier; however, we can tolerate a poor‘vapor barrier. The fact that we locate the gas barrier on the cold side of the wall precludes the possibility of having one membrane serve both as a gas barrier and as 'a vapor barrier. The combination of a gas barrier on the inside and a vapor barrier on the outside form a wall structure from which moisture escapes with difficulty. The vapor barrier is usually more permeable than the gas barrier; therefore, the gas barrier functions to retard the movement of moisture out of the wall. Vapor enters the wall at a much larger rate than it can escape from the wall and the accumulation of moisture takes place inside the wall usually close to the gas barrier; in some cases it may migrate to and condense on the back side of a metal gas barrier. It is necessary that the gas seal be accessible for maintenance and testing. Because of this requirement, the most practical loca- tion of the gas seal is at the interior surface of the storage room. When a gas barrier is present at the inside of a wall structure, the vapor can no longer pass through the wall. If the gas barrier is at the cold side of the wall, the pressure differential is positive. The moisture will diffuse into the wall until it reaches a point of storage, the maximum penetration being the gas barrier. If the gas barrier 138 side of the wall becomes the warm side, the pressure differential is negative and a diffusion process will transfer moisture out of the wall. The moisture which is diffused out of the storage room wall will usually be stored as hygroscopic moisture until the wall materials reach the saturation point. After saturation is reached, the moisture will run off as free water. A CA storage wall with the gas barrier on the inside can be designed to facilitate a favorable moisture balance within the wall, according to Lorenzen (1963). Some Of the moisture accumulating problems of a CA storage wall would be eliminated if the gas barrier could be located on the warm side of the wall, the greater part of the year, where a single barrier would serve both as a gas barrier and a vapor barrier. 3E. Analysis of Design Principles and Possibilities 3E1. One direction vapor flow When the vapor is flowing in the structure only in one direction because of favorable weather and operating conditions, the basic design principle is to avoid vapor condensation within the structure; this is the flow-through principle. The flow- through principle (BRAB 196 3) " moisture which flows into the insulated system should not be allowed to become trapped therein, but should continue to flow completely through the system. " Several commercial designed pro- cedures are based on this principle. For example: 139 1. Fail safeconcept of Lotz (1964). A fail safe system is one that provides a highly efficient vapor barrier on the warm side with a porous insulation and a porous interior finish. This allows any vapor that does penetrate the vapor barrierto pass through the wall to the inside without any intermediate vapor dams to cause condensation within the insulation system- 2. The mechanical system. This system is based on the principles that (a) the best vapor barrier economically available is used on the warm side; (b) an insulation with a moderate water vapor permeability (4 perm per in.) plus a relatively low air permeability is desirable; (c) a cold side finish that is highly permeable to vapor; (d) eliminate vapor dams on the cold side of the vapor barrier. McGarvey (1964) gives a detailed description of the materials and construction procedure. An adoption of these methods for CA construction would require that the gas and vapor barrier both will be the same barrier and located on the warm side. 3. Dow Chemical method. The Dow Chemical method uses exterior aluminum finish as the gas and vapor barrier on the warm Side (Dow 196 5) . 140 3E2. Reverse flow When reverse flow conditions are expected to exist in a wall for extended periods of time, then the economics, temperatures, dew points, and the duration of the reverse flow cycles should be studied to determine which one of the following possible practical solutions to the problem of moisture accumulations in the wall will be chosen. Possible Reverse Flow Solutions l. Vapor barrier and gas seal on one side of the insulation system. Unless the vapor pressure gradient reversal is particularly severe, a single vapor barrier and gas seal may be located to re- strict vapor movement under the most severe conditions with some condensation being tolerated when the vapor pressure reversal occurs (Lotz 1964, Hutcheon 1958). The Dow Chemical Co. is using 15 mil embossed aluminum with a baked on enamel as the outside covering for their new design of insulated panel for CA apple storage con- struction in Michigan (Dow 1965). The aluminum skin on the outside of the panel plays a triple role--it provides the weatherability of the side walls themselves, and serves as the vapor seal and gas seal. The panels are constructed around Styrofoam insulation which is sandwiched between sheets of 1/4 in. thick exterior grade plywood. 2. Seal both sides of non- permeable insulation. Theoretically it is a good solution to seal both sides of a non-permeable insulation, 141 one vapor barrier on the warm side, and one vapor barrier on the cold side. Since the commercial installation of vapor barriers is usually relatively poor, and the warm side vapor barrier is more difficult usually to apply than the cold side, in the practical end, a poor warm side barrier is installed with a moderately good cold side of the barrier. In this case, the cold side barrier serves as a vapor trap and maintains and holds the vapor that moves into the wall preventing its escape and allowing it to be present because of de- terioration of the wall. Only when the insulation is hermetically sealed in the wall it is a good solution. However, the high cost coupled with severe technical fabrication problems make this solution undesirable as far as controlled atmosphere storages are concerned. 3. Install a vapor andgas barrier in the insulation with ade- miate insulation on each side of the vapor barrier to prevent condensation. There is a design trend to use ventilation to prevent conden- sation of water within the enclosed wall. There are several different methods using ventilation to prevent condensation. These methods are analyzed below: 4. Lorenzen's (196 3LbalanciAgtechnique for moisture control. Three factors are considered for vapor balance within the CA storage walls. (a) The environment within and outside the storage and its 142. wetting and drying effect on the wall. (b) Potential moisture capacity of the wall. (c) The movement of moisture into and out of the wall and the resistance to the movement of moisture into and out of by varying materials and types of construction. The amount of vapor which may be allowed to be transmitted into the wall for storage within the wall is limited by the moisture storage capacity of the wall. MAtAP = [H1(m'l' - m'1)+ Hz(m3 - m'z)+ ' - °]7000 (3E-l) The vapor pressure of the ambient air at the end of the drying period will determine the maximum equilibrium moisture content available for a given hygroscopic material. The safe moisture content should be realized with a vapor pressure just below the dew point vapor pressure for the ambient air temperature with a relative humidity of about 95 percent. The hygroscopic moisture storage potential of the wall will be exceeded if a vapor barrier with too great a permeability is installed. If a vapor barrier with too great a resistance to vapor flow is installed, the full drying potential of the drying period will not be realized. When the full natural drying potential of the wall is not realized, mechanical drying of the wall may become a necessity. Generally, if the area of drying potential (when AP is negative) exceeds 143 the area of wetting potential (Fig. 3C-1), a dry wall may be main- tained naturally. When wetting potential exceeds the drying potential, then a wall compartment ventilation system becomes a necessity (Fig. 3E-1). This ventilation system must be designed to take ad- vantage of any periodic drying potential. Automatic controls would operate when actual vapor pressure reversals occur; these vapor pressure changes would be sensed by dew point indicators. A ther- mostat set at the controlled atmosphere operating temperature could also be used to control the system. When a drying potential exists, dampers will open and ventilation will start. When a wetting poten- tial exists, dampers will be closed and ventilation will stOp. The capacity of the ventilating system will be determined by the rate and quantity of air movement required to achieve a favorable drying effect. Vapor will pass through the insulation and condense on a surface, such as the back of the gas barrier which should be separated from the insulation and major framing material so that condensed moisture will drain away without saturating these materials. The summer non-operational period is a period of drying potential which has to be taken into consideration. The length of this potential drying period will vary. It can be extended when the room is put into operation later in the fall and also when the room is opened and the fruit is removed earlier in the spring (Fig. 3C-1). The rate of vapor transmission into and out of the wall can be 144 Amoofi :LNCLLCJC m“-_.3a_AVAna~nAc.. H...~_z_.3~3._~mw .omnzw._ac~mu m\.mv F_L~L~._>> ~5o.n_finac.. .o.__z~mw_xc~__ ..Acm ~_Mu_mw.onv ..Ao mw.o.__5~...o.m . _ - m_m, .wa..; To” 3. u_ F l b m2. ILHWWH, o_=m , _=m \. :mm 2)) r. Om; O n B B.§ 5 § . Nflb .Edmo wk .......... amazqo m. ZOCdJDmZ_ JOKFZOU 44mm mdo mudido oz< 24¢ PmD i i I); Water Fig. 48-2. Relief \’.‘:1\'O to prexent excessive pressure in CA room {Smock and Blanpied 1960) 158 operates in parallel with the leaks in the storage room; therefore, to be effective, it must apply and recover air with very small pressure differentials. A breather bag acts to reduce air leakage due to temp- erature cycling and at the same time acts as a safety device for the room. In general, a properly designed breather bag may reduce the leakage from temperature cycle by 50 percent. However, the breather bag will not reduce leakage from pressure differentials induced by fans, winds or changing barometric conditions. The breather bag provides a good guide to the pressure condition in the room. A con- stantly deflated bag indicates that air is being continuously pumped from the room whereas the constantly inflated bag suggests air is constantly being added to the room. Oxygen gain due to carbon dioxide removal is also due to the control of the breather bag since the volume of CO; scrubbed out of the room must be replaced by air from outside the storage room before an equilibrium can be reached. The breather bag will improve the operation of a tight room but will do little for a leaky room. The greatest benefit of the breather bags is in small rooms which tend to be subject to rapid changes in temperature and often have a large space factor. 4C. Measuringghe Gas T:igh_tness of CA Storage Rooms Two general methods of measuring the gas tightness of CA storage rooms have been used and those may be appropriately referred 159 to as rate of leakage tests and the pressure drop test. Rate of leakggg test. In the rate of leakage test method of CA storage rooms, air is metered into the room at a constant rate, for example, 25 cfh, until the differential pressure across the wall reaches a constant value. At this point, air is leaking out of the room at the same rate as it is being added to the room. Therefore, the manometer reading is the differential pressure producing 25 cfh of leakage. Following the initial determination, the rate of flow is increased and the entire procedure repeated for a second flow rate. The final result is a graph relating pressure differential with rate of leakage. Rate of leakage type tests are important in externally gener- ated gas systems where a certain amount of atmosphere must be exhausted from the room per hour. However, in fruit generated storage rooms, the rate of leakage test has limited practical value because the normal pressures that the fruit generated CA room operates under are so low that micro- manometers are necessary to measure the pressure differentials. If the rate of leakage type test is carried out, variables such as changing temperature and baro- metric conditions can cause greater effects on the pressure than the air flow rate being evaluated. Theoretically, rate of, leakage tests can provide a great deal of information on the behavior of the CA storage room. For practical 160 purposes the rate of leakage tests require too long a time and too complicated equipment for general use in CA storage room evaluation. Pressure difference CA storage room test. In the pressure difference CA storage room test, the room is sealed and the pressure inside the CA storage room is raised to 1 in. of water gauge by means of a small centrifugal fan or pressure type vacuum cleaner. When the desired pressure of l in. of water gauge has been reached, the fan is turned off, the room sealed and the time recorded and pressure differential recorded at 10 or 15 min. intervals during the next 1-2 hours. If a pressure of O. 1 in. of water gauge remains after 25 min., the rooms in most cases are sufficiently tight to operate well as CA storage rooms. In general, for a given pressure test result, the performance of the room will vary directly with the size in that large rooms will perform well and small rooms will perform less well. In general, rooms of less than 10, 000 bu. capacity should require 60 min. for the pressure to drop from 1 in. of water gauge to O. 1 in. of water gauge. SECTION 5 PRECOOLING AND AIR DISTRIBUTION IN A CA APPLE STORAGE ROOM 5A. Introduction The present day designs of refrigerated storage rooms using, in some cases, ceiling mounted evaporators with propellor fans, in other cases, floor mounted evaporators with centrifugal fans, some- times with ducts, sometimes without ducts, may appear to be hap- hazard; however, a great deal of time and energy has gone into attempts to improve the general performance of the storage room. The product is usually brought in from the orchard in pallet boxes and these boxes are moved into the storage directly from the loading dock. The boxes are stacked neatly in the storage room; in some cases by design, space is left between the boxes; in other operations again by design, the boxes are stacked as closely as possible. In all storage room designs, the objective is to have uniform air flow throughout the room and a uniform temperature throughout the room. Unfortunately, in large storage rooms it is extremely doubtful if these objectives are ever met. When boxes are stacked in a large room and the air is rather 161 162 grossly circulated throughout the room, the problem is indeed quite complicated. Some of the complicating factors stem from practical situations; for example, the cold air is normally distributed at the top of the room, the cold air will normally move toward the floor of the room due to its density; however, air, as it flows through the stack and becomes warm, tends to rise which in some circumstances we may find that the natural convection forces are offset by the den- sity effect of the mechanical circulating system. As the point in the room is farther from the evaporator discharge, evaporator inlet or duct discharge, the area is more subject to natural forces than to the mechanical forces; whereas then near the evaporator discharge or duct outlet or near the evaporator inlet the flow pattern depends more strongly on these mechanical effects and the forces generated by these devices. Many researchers have studied the flow paths of air in specific types of rooms and under special types of loading and operating conditions. The type of box arrangement has been studied extensively and specific recommendations were presented by Sainsbury (1961, 1962) regarding the distance between rows of boxes and between the last row of boxes and the wall of the storage room. Gray and Smock (1955) have studied the air distribution from ducts. Hukill and Smith (1946) studied air distribution and precooling rates and designed a reverse flow system that permitted lower air temperatures and faster 163 cooling in a rather unique system. In all of these systems there has been a tendency to develop the system around practical guide posts. Outlined below is a method of operation designed to obtain maximum use of the natural forces and supplemental mechanical forces. 5B. General Des_ign of New Method of Storage_ In this new approach to the problem of the storage room, its air distribution and precooling rates, all of the spaces between the pallet boxes will be closed and the stacking will be arranged in such a way that the air will be forced to flow only through the void spaces between the apples in the pallet boxes. In this way it will be possible to obtain a maximum cooling surface. After the air has moved up- ward through the stack of apples, it will move along the ceiling and wall, after which it will be cooled, allowed to move down toward the floor of the storage room where it can again move upward through a stack of product as shown in the schematic drawings in Fig. 5B-1 and Fig. 5B-2. Suction fans distributed uniformly on the false ceiling will be used to move the air through the system and give even distri- bution in the stack. To make possible the upward flow of air through the stacks of pallet boxes, a specially designed pallet box will be required in which the bottom has about 50 percent voids (about the same as in a pallet box full with apples). To prevent air from moving horizontally through the channel form by the stringers under the pallet box, an extra bottom or pallet (with 50 percent voids) can be 164 SUCTION FANS CCOLING COlLS ‘_\ ‘ \ CEILING CLO? H S(‘hvmntie drawing of "packed bed" room- - side Vli'VV I912; 5H-1. H 1 O Q FAN \ A... COOLING \/ V'CEILRNG COILS \ Fig. vii-2. Sehemntie dr-iwing of "parked-bed" room--top View 165 placed under alternate stacks so that the pallet openings are stag- gered, as shown in Figs. 5B-1 and 5B-3. To prevent air from moving back into the top of the boxes, a cloth baffleis hung vertically over the front of the boxes as shown in Fig. 5B- 1. This "packed bed" approach, can be efficiently used for the prevention of short cir- cuited air flow during precooling during the loading period (Fig. 5B-4). Using only those suction fans in the area in the room where the room is loaded plus the placing of baffles in strategic locations will force the air to move up the stack of fruit as shown in Fig. 5B-5. To pre- vent channeling in the stack itself, the fans can be used alternately for drawing and blowing the air. 5C. General Outline for Analytically Determining the Value of the Variables in the " Packed Bed" Type Storgge Room OperatingSystem In developing this analysis it is assumed that there is one con- tinuous box of apples extending from the floor of the room to the ceiling. Or, in the chemical and mechanical engineering terminology, a "packed bed" of spheres, where in practical sense the spheres will be apples or other food products. In considering the flow of air through this packed bed, the initial design criteria will be the temp- erature and air velocity through the bed to give the desired cooling rate. Once the desired velocity through the stack has been determined, it is considered that the apples in the box will cool at the same rate i: . j/ LL/ L. : i ...EXTRA PALLET w I // Fig. 58-3. The use of an extra pallet every second row to prevent horizontal air Channeling i r , i | F L I Fig. 58-4. Short Circuit air flow in regular room precooling during loading period FALSE /CE|L|NG Fig. 513-5. Precooling during loading—"packed-bed" approach 167 as an individual apple exposed to the same air velocity. This will allow the use of the experimental data available from single fruit cooling experiments of Kopelmangtgl. (1964). Using the "packed bed" principles, the pressure differential necessary to produce the desired air velocity through the boxes of apples will be calculated. The quantity of air necessary will be the product of the void area times the velocity. Once thequantity of air required and the pressure differential has been determined, the fans can be sized directly. Knowing the cooling rate data, the quantity of heat to be removed, the refrigeration capacity can be determined as soon as the heat re- moval rate has been evaluated. 5D. The Method To Calculate Design_Data for the "Packed Bed" Room--Precoolirg and Air Distribution The best conditions for the. "packed bed" cooling- air distribution problem will be calculated now, involving the following variables: t time T temperature of cooling media T initial temperature of the apple T temperature of the apple h film coefficient v air velocity Ap needed to get a velocity v Q heat to be removed 168 Transient state cooling of spheres, in a packed bed arrangement, generating heat (respiration heat which is a function of the temperature of the apple), under a constant flow of cold air, is assumed. As the air is circulating constant air temperature is assumed. 5D]. The approximate calculation of T1 and h To establish the exact cooling curve (T versus t) the first approximation of T1 and h must be known. The easiest way to ap- proximate h and T1 is to assume cooling of a sphere without genera- tion of heat using the method suggested by Pflug and Blaisdell (1963). The procedure for this approximation will be as follows: The relationship between the f value and the air temperature is given by Eq. 5D1- 1. 1g (T ‘ T1) : t/f + 10g j(To ' T1) (5D1'1) Ta - T1 3: _To _ T1 (5D1-2) Assuming that it is desired to cool the apple in t hours from To to T there are three unknowns T1, j and f. But if f is assumed NBi is found in Fig. 2 of Pfluggtgl. (1965), and j (Jc, Jm or J5 as needed) in Fig. 3, 4 and 5 of Pflug§t_a_l_. (1965). h is calculated from NBi and then T1 is calculated from Eq. 5D1-1. 169 5D2. Calculation of the "exact" cooling curve Assumptions and boundary conditions 1. It does not have to- be assumed that the whole sphere is at the same initial temperature, a more realistic approach is that the temperature distribution in the sphere assuming equilibrium conditions is: 2 2 __ z a " r _ T—TOS+ b ( 6 ) (5D21) QR where b‘2 = (5D2-2) PCS 2. The sphere loses heat to its surroundings at a rate propor- tional to the temperature excess above the surface. 3. Heat flow is uniform so that the temperature is a function of the radial coordinate r, and independent of angular coordinates. The fundamental cooling equation for a sphere generating heat (con- stant QR) is: d H -313. 29:]; 2 _ -rzdr(r dr)+b (5D2 3) Q. 1'. taking a new variable u = rT this becomes — = bzr + — (5132-4) The solution of the differential equation with the assumptions and boundary conditions mentioned (Awbery 1927) is: 170 +LaZ-r2)b2 kabz 1 e-¢Zat/a2 . 9;: T 2 T1 60 + 3ha + ; 2 D s1n a (5D2-5) l _ Zath [4»sz + (ha - maf- Where D — d) ¢2k2 + (ha _ k) (5DZ 6) F-(T -T)-'1S§—b: (5D2-7) _ OS 1 3ha tan d; = JL (5D2-8) k - ha The sphere temperature at any desired point within the sphere is a function of the heat of respiration, surface heat transfer coeffi- cient, initial surface temperature, temperature of the surrounding air and time (Eq. 5D2-5). However, it has to be kept in mind that this was calculated for a constant QR. In order to overcome the problem that QR is changing with the temperature a numerical method will be applied at this point. The relationship of the heat of respiration with apple temperature based on the data1 in (U. S. D. A. 1954) for the Jonathan variety was found to be, for the second polinomial degree (T = ° F). QR = 3. 59'1“2 - 248T + 5024 (5D2-9) Program E2 UTEX LSCFWOD-- " Least square curve fitting with orthogonal polynomials" of the CDC 3600 computer of MSU was used to calculate Eq. 5D2-9. At this stage a finite difference numerical 1The data are not enough therefore this equation is only approximate. 171 analysis will be applied. It will be assumed that during a very short t_i_n;1§ QR is constant and the temperature T of the sphere can be found by Eq. 5D2-5. The new T will be used now to calculate a new QR with Eq. 5D2-9 and the new QR to calculate new T, (etc. The error can be controlled by adjusting the At. In this way the relationship T versus t can be found. It is obvious that the last step can be most efficiently calculated using a computer. To get accurate results At can be made as small as needed and as many roots for the tan ¢ as needed can be used. The advantage of this numerical solution is that it will be easy to use the same procedure (program), change the variables; sphere size, initial apple temperature, air temperature, h and time, to get information about the effect of each variable and get the valuable design information needed for any specific case. It is now clear that the h and T1 found by Pflug's method (Eq. 5D1-1), ignoring the heat generated, have not influenced the results based on the "exact" method to evaluate T versus t. If t is found to be too long T1 or h will be changed in order to obtain the desired T. With the "exact" method the theoretical T versus t will be obtained and the use of f, j as with Pflug's method (Eq. 5D1- 1) is not needed. It should be kept in mind that Pflug's method was developed with the assumption that when the time function is large, the curve 172 becomes coincident with the straight line asymptote. However, in this case, cooling of apples in rooms, the cooling process is rela- tively long and the effect of the heat of respiration is especially im- portant at the beginning where the temperature of the apple is high. As Pflug's method is much easier to use it may be possible to find a correlation between the two methods. In other words, it might be possible to find apparent f,(f') that will include the effect of heat of respiration. As the apple is cooledAT'f becomes smaller and smaller and Q approaches a constant value therefore, t will have to be definite in the program. As t —> 00, T approaches equilibrium temperature where QR = constant, and the surface temperature exceeds the temp- 2 erature of the surrounding by This is the steady state stage 3ha ' or the regular storage period. 5D3. The determination of the air velocity v needed and chm A large amount of experimental information on heat and mass transfer in packed bed has been analyzed to arrive at the following empirical correlations (Bird e_t_a_l. , page 411, 1965). 0.91 Re-O' 51 it) (when Re< 50) (5D3-1) JH -0. 41 0.61 Re ti: (when Re>50) (5D3-2) JH Here the Colburn factor (jH) and the Reynolds number (Re) are defined by 173. C_|.L 2/3 . h ( PA J :——C_ -——— (5D3-3) H PAGo k GO V Re = = —p—— (5D3-4) 33% q” as“? In this equation the subscript f denotes properties evaluated at the "film temperature" Tf = % (TS + T1) (5D3- 5) LP is an empirical coefficient that depends on the particle shape. For a sphere a value of LP = l was found. Since the desired h is known (calculated in Sections 5D1 and 5D2), J is found from Eq. H 5D3-3. The Re is calculatedusing Eq. 5D3-1, or 5D3-2, depending on the range of Re numbers. v is then found from Eq. 5D3-4. Knowing v, the quantity of air to be circulated can be found from Eq. 5D3-6. chm = vAe (5D3-6) where A. is the cross section area of the stack and e the void fraction. 5D4. The determination of maximum APB needed for air distribution Using the "packed bed" approach (Bird e_tgl. , page 196, 1965) the APB (needed for the desired v can be calculated. The correlations to calculate APB are based on data from experiments with perfect spheres which were smaller than an average apple, therefore, the present procedure can only give an 174 approximation or trend and will have to be checked thoroughly experimentally. The general Ergun (1952) equation which is suitable for a large range of Re numbers is suggested to determine the APB AP p 3 B 2a 6 (l-e) — _ : + . " G02 (H) (1_€) 1501;13:0/H 175 (5D41) where G0 = p€V Preliminary calculations of AP using Eq. 5D4-1 showed that a very B low APB is required: to obtain velocities of 5- 100 f. p. m. the pres- sure drop AP , was found to be .001-0. 3 in. water gage. B The air flowing along the ceiling, walls and floor is subject to friction forces; the fans have to be designed to overcome this friction in addition to the friction of the flow through the apple boxes. This difference in pressure AP will be calculated assuming air flow D through a duct with a rough surface. Lentz and Nakano (1961) de- scribed a method of predicting air friction pressure losses in shallow ducts. This method can be used for our purposes. The fan has to be designed for a total pressure drop, APT, of : A .. APT APB+ PD . (5D4 2) 175 5D5. Evaluation of the total refrigeration load The maximum refrigeration load will be calculated for the first day of precooling (full room) when the temperature drop and the heat of respiration are largest. The total general heat balance is: sensible heat heat of respiration To MCPS(T0 — Td) + M (3. 5gTz - 248T + 5024) dT r-iR d (5D5-1) + Q C (T1 ' T2) fans + Qleakage : chm PA The sensible heat, heat of respiration, heat produced by the fans and heat leakage into the room, are removed by the air at temperature T1 and velocity v. The heat of respiration is calculated by integrating Eq. 5D2-9 for one day. The refrigerant temperature, T3, can be found using Eq. 5D5-1. Another important calculation will be the calculation of the surface temperature of the apple. By substituting r = a in the "exact" method, the surface temperature versus time will be found. This is important because the air temperature can be lowered as long as there is no danger of freezing the apple and during this period of time to obtain faster cooling with the same APT and v. Advantage s 176 SE. Advantages, Disadvantages and Discussion of the New App_roach 1. As there are no spaces between the boxes more apples can be stored in the same size room, therefore: (a) (b) There is a possible saving on storage construction costs per bushel ft.3 . Space factor 2 (TD-11') (Sec. 2D1), IS smaller (the amount of air (oxygen) in the room is smaller), therefore, the time to reduce the oxygen level in a regular CA room is shorter and the gas consumption using burner units to reduce the oxygen is smaller. 2. Since the air circulation is based on the minimum air movement; (a) (b) (C) Infiltration will be reduced because of reduced air velocities. Less heat is added by the fans, therefore, the refriger- ation load will be smaller. Smaller fans will be required and the energy consump- tion will be reduced. 3. Because of the unique air distribution: (a) (b) A promising possibility for even distribution of air. Excellent air distribution where precooling of room during loading is needed. 177 Disadvantagfi 1. The need for standard size pallet boxes and rooms and especially designed pallet boxes. 2. Stacking of boxes in the room is more complicated. 3. There is a danger of short circuiting air in case of poor stacking. 4. Insulation protection is not adequate because curls and bumper rails will interfere with the required spacing of the boxes. This study was based on theoretical analysis without any experimental work and at this stage it is only a speculation of a new air distribution method. A general outline was given on how to calculate the various design factors for a storageroom. A. com- plete program to solve this complicated problem was beyond the scope of this thesis and results of experimental work to prove the advantages of this method are needed. SECTION 6 DISCUSSION AND CONCLUSIONS Controlled atmosphere (CA) storage is a method of keeping living fruit longer by refrigeration, reduced oxygen and higher than normal carbon dioxide concentration in a gas tight insulated room. There are a large number of construction materials available in use in the specific area of CA storage construction. The most common type of construction scheme in use is that of using concrete block or tilt up concrete with board form insulation, or prefabricated insulated panels. In most of the CA rooms today galvanized sheet steel (28- gage) on the inside is used to provide the gas seal, and vapor barrier at the outside surface of the building is used to prevent vapor trans- mission into the area. The room has to be gas tight and the gas tightness is a very critical factor determining the satisfactory operation of CA room as will be discussed below. There are a number of alternative systems to control the atmosphere in the CA room. The simplest one is restricted ventila- tion in which the sum of the oxygen and carbon dioxide concentrations is always 21%, but that does not meet our present need of low oxygen and carbon dioxide. 178 179 The control of the atmosphere to achieve the desired CO; and 0; levels is done in either of two ways. 1. By removal of the excess CO; by physical absorption (water), chemical absorption (NaOH, KZCO3, MEA. or dry lime) and by using molecular sieves, and reduction of oxygen in the room by the 5... fruit itself, addition of liquid or gas N; or COZ, or by burning the oxygen in the room atmosphere. 2. By external gas generator. The control of the atmosphere is made by introducing an externally generated atmosphere into the room. As there are many alternatives to reduce the CO; and 02, many different combinations of carbon dioxide scrubbers and oxygen reduc- tion systems can be used to control the atmosphere. In order to choose the best method for control of the atmosphere, some critical factors have to be considered: 1. Degree ofgas tiLhtness in the CA room. If the room is very tight every combination of carbon dioxide scrubbing and oxygen reduction system can be used. In this case precautions against pos- sible damage to the room because of pressure changes have to be considered. The possible damage can be reduced by eliminating the cause of pressure cycles or incorporating a safety device that will prevent the build- up of excessive pressure in the room. If the room is not very tig_h_t (0. 5 air-changes per day or more), combination of 180 water scrubbing and natural respiration cannot be used and the use of nitrogen flushing and liquid nitrogen are not desirable. Although there are methods to control the atmosphere in a leaky room using an externally generated atmosphere or combinations of commercial burners with scrubbers on a recirculatory basis, the operation of a tight room is easier and more effective. The time required to reduce the oxygen level in the room is shorter and the gas consumption using a burner to reduce the oxygen level is smaller than with a leaky room. Therefore, it is recommended that when building a CA en- closure it be made as tight as possible. However, in case a severe gas leakage develops because of failure in building construction or sealing, a combination of burner to reduce the oxygen with a CO; scrubber, or external gas generator is needed. 2. Management practices. It is necessary to know if the CA storage room is designed for single or multiple opening during sea- son. Sing_l_e opening. --If it will not be necessary to re- open the room during the season, only one pulldown of oxygen will be needed (ex- cept for emergency). In this type of operation any combination of carbon dioxide scrubber and oxygen reduction system as mentioned before can be used. For such an operation, dry lime and water scrubbing were found to be the most economical carbon dioxide re- moval methods. A combination of liquid nitrogen for flushing and cooling appears to be potentially the most economical way to reduce 181 the oxygen level in the room and the burners comprised the next most economical group. 3. Multiple opening. If multiple opening of the rooms during the season is desired, practically all the combinations of scrubbing with natural respiration are not desired. Therefore, only combina- ‘F‘ tions of scrubbers, with burners and liquid nitrogen are desired. The operation of a partially filled room will eliminate all the combinations of scrubbing with natural respiration and nitrogen gas or liquid (be- E _ cause too much liquid nitrogen is needed to operate a partially filled room economically). Only carbon dioxide scrubber combinations with burners are suitable. For such an operation, dry lime is the most economical carbon dioxide removal method and the burners are the most economical oxygen reduction systems. Comparison of annual costs of carbon dioxide removal methods, oxygen removal methods, various combination of carbon dioxide scrubbing and oxygen reduction methods and commercial complete methods to control the atmosphere, are given in Tables ZJ- 3, 4, 5 and 6, for different volumes of storage utilities and labor prices. The comparison of the results inTables ZJ- 3, 4, 5 and 6 shows that combinations of scrubbing and oxygen reduction systems for a specific volume and utilities price level might be more economical than using one of the complete systems. As a typical example a combination of an Arcat burner with a water or dry lime scrubber for 182 60, 000 bu. is more economical than the most economical complete system for this volume of fruit (which for both large and small operations is the Arcagen). In a small scale operation the annual cost of water and dry lime scrubbing (based on 3 cents per bu. construction cost) is similar. In 11:“- a large scale operation the annual cost of water and dry lime scrubbing '. (based on 1.5 cents per bu. construction cost) is also similar. There- fore, the choice between water and dry lime scrubbing or a possible % change from water to dry lime scrubbing should be based on local ' conditions and prices. The advantage of water scrubbing for main- taining high relative humidity should also be taken into account. For operation of filled and tight rooms and especially rooms which are not opened during the season, the combination of flushing and cooling or flushing only with liquid nitrogen shows promise. How- ever, this combination is in the experimental stage and depends on the liquid nitrogen cost. The use of liquid nitrogen for maintaining the atmosphere during the season is much more expensive. The Eaves "Oz controller" was used on a small scale only, but is very promising economically and is worth looking into for a larger scale system. The use of Arcagen is economical for 60, 000 bu. or multiples of 60, 000 bu. An introduction of a smaller unit may be economical for a smaller scale system. The use of a Tectrol gen- erator on a purchase basis ($5000 per generator for 20, 000 bu. was 183 assumed) is much more economical than on a lease basis. Every combination of carbon dioxide scrubber and oxygen reduction device is suitable as a complete system for the control of the atmosphere. In order to choose the most suitable system, some preliminary elimination of these methods can be made easily accord- ing to the degree of gas tightness of the room and the management practices as mentioned before. After this elimination step, the most economical combination among the remaining can be chosen from Tables 2J- 3, 4, 5, and 6 (which are cost data for CA unit operation) according to the volume of storage and utilities and labor prices of the particular case. The problem of carbon dioxide and oxygen measurement in a CA room is whether to use Orsat or a combination of carbon dioxide and oxygen instrumental analyzers and recorders and also, which size storage (how many rooms or measurements) will justify the use of expensive instrumentation. An approximate analysis based on comparison of annual costs of operation of Orsat and the instrumentation system showed that with more than 7 storage rooms, a commercial instrumentation system is more economical than using Orsat. When a relatively simple instru- mental system based on "local" design and installation is used, it was shown that with more than 4 rooms this sytem is justified. The mechanical refrigeration precooling of apples in the room 184 is a rather expensive method and an improvement of the air distribu- tion method in the room during precooling is still needed. Recently liquid nitrogen was tried experimentally to serve both for partial cooling and for flushing the oxygen from the room. The use of liquid nitrogen is advantageous because there is a peak of refrigeration load at the first few days of the storage season when the field heat has to be removed. The refrigeration load calculation for sizing a mechani- cal refrigeration system is based on the maximum load during the precooling time. However, the precooling may be also done by a combination of a liquid N2 and smaller mechanical refrigeration system. The mechanical system will be designed to hold the apples at the desired temperature after the precooling period _(_)_n_ly. Approximate calculation for precooling apples from 65 to 35° F with mechanical refrigeration showed that the annual operative costs per year were 4. 18 cents per bu. based on 1 cent/ kwh and 5.2 cents per bu. based on 3 cents/kwy. Precooling (65- 35° F) and flushing using liquid N; gave costs of 3. 5 cents/bu. when the price of liquid NZ was 54 cents/100 cu. ft. N2 gas and 2.4 cents/bu. when liquid N; was 37.5 cents/100 cu. ft. N2 gas, which shows that liquid N; pre- cooling and flushing is more economical than using mechanical refrigeration only to precool. The use of liquid N; was only checked experimentally on small scale basis (1500 bu.) and has to be checked on a larger scale. It is too expensive to control the temperature and 185 the gas levels in the room for the whole season and especially when the room is not filled with apples, but is a potentially promising solu- tion for precooling and oxygen reduction when only a single opening of the room is needed. Another approach to precooling and air distribution in a CA storage room has also been presented. In this new approach, all the spaces between the pallet boxes will be closed and the stacking will be arranged in such a way that the air will be forced to flow only through the void spaces between the apples in the pallet boxes. In this way it will be possible to obtain a maximum cooling surface. After the. air has moved upward through the stack of apples, it will move along the ceiling and wall after which it will be cooled, allowed to move down toward the floor and then again upward through the stack. Suc- tion fans distributed uniformly on the false ceiling will be used to move the air through the stack. The most important advantage of this method to CA operation is that there are no spaces between the boxes in the room and more apples may be stored in the same size room. There is a possible saving on storage construction costs per bu. and the space factor Z (cu. ft. /bu.) is smaller, 1. e. the quantity of air (oxygen) in the room is lower. The time to reduce the oxygen level in the room is shorter than in a regular CA room and the gas con- sumption using burner units to reduce the oxygen is lower. This approach is only a speculation of a new air distribution method. APPENDICES APPENDIX 2-1 Comparison of Annual Cost of Operation of Orsat and the Instrumentation System Described in Fig. 2G-l Annual measurement costs--one room per season usi_ng Orsat Initial cost of Orsat: $85 Service life: 3 years Measurement of one room: 15 min. Wages: $1. 5/hr. Season: 200 days CRF: capital recovery factor (see Sec. 2J) (3 years 5 percent interest): 0. 36721 Operation cost per season: (:75) X 1. 5 X 200 = $75 Annual cost P (CRF) + operation cost 85 X 0.36721+ 75 = $106 per room per season Approximate costs for the instrumentation based on "local‘l desigp and assembly work flip to 15 rooms) Assumptions: Service life: 10 years 186 187 CRF (10 years, 5 percent interest): 0. 12950 Operation time: 0 Calibration time will be the same as Orsat checking and repair therefore will be ignored. For the whole system Beckman Model 778 - oxygen analyzer $ 475 Beckman Model 7C Thermal Conductivity Analyzer (CO; analyzer) 890 Two pen recorder 1300 Sampling devices 110 $2775 For each room sampling and switching system--(time switch, sole- noids, piping valves installation, etc.) where n is number of rooms: $150n Total instrumentation cost for measurement and recording: $2775+ $150n Therefore: Orsat operation costs Total instrumentation Operation measurement and recording costs lO6n : (2775 + 150 n) + O n = 4 The additional costs of control of liquid nitrogen for example will be $20 + $45n therefore the total instrumentation cost for measurement, 188 recording and control will be approximately (2775 + 150n) + (20+ 45n) = 2795 + l95n dollars Cost of commercial instrumentation system based on proposal from Beckman Instruments, Inc. (1965) Assumptions: Service life: 10 years CRF (10 years 5 percent interest): 0. 12950 Operating time: 0 Cost of system For For 15 rooms 7 rooms Beckman 778 oxygen analyzer $ 475 $ 475 Beckman 7C thermal conductivity analyzer $ 890 $ 890 (carbon dioxide analyzer) Automatic stream selector $ 925 $ 633 Sample handling system, unassembled $ 1570 $ 1170 Panel including mounting of instruments $ 1265 $ 1265 Baily two pen circular chart recorder $1.399 113%) Total $6435 $5733 Therefore, based on 15 room system cost, 189 Orsat operation 'l‘otal instrumentation measurement Operation. costs and recording for 15 rooms cost 106n = 6435 X 0.1295 + 0 n = 8 There is some saving on automatic stream selector for smaller than 15 number of rooms (automatic stream selector 10 points - $633) and on the sample handling system (for one room less than 15 subtract $50 per .room from $1570). Therefore, the total cost of a system for 7 rooms will be as shown above and based on 7 room system cost 106n 5737 X 0.1295 + 0 n=7 With more than 7 rooms the commercial system is more economical than Orsat. APPENDIX 2 - 2 CO; Removal Costs Arcosorb for 60, 000 bu. Initial cost: $9500 Shipment to Michigan: $ 100 Installation: $ 350 Years in service: 10 Salvage value: 10% Operational costs Power costs 1 cent/ kwh: $220 Power costs 3 cents/kwh: $660 Water: $14. 3 Labor, 8 hr. per room per season $1. 5/hr. : $36 Taxes and insurance (2%): $190 K;CO3 - Sulzer Brothers Winterhur Switzerland Initial suggested retail cost in Europe for 400 ton (19,500 bu.): $5,650 Years in service: 10 Salvage value: 10% 190 191 Operational costs Power 1 cent/ kwh 20, 000 bu.: $ 51. 7 60,000 bu.: $155.1 3 cents/ kwh 20, 000 bu.: $155.1 60, 000 bu.: $465. 3 Labor, 8 hr. /room/ season when $1. 5/hr. 20,000 bu.: $12 60, 000 bu. : $ 36 Taxes and insurance (2%) 20,000 bu.: $113 60, 000 bu. : $ 339 Water Initial cost based on $1400 for 20, 000 bu. unit \Years in service: 10 Salvage value: 10% Operational costs Power 1 cent/ kwh: $ 35. 8 per 20, 000 bu. 3 cents/ kwh: $107.0 per 20, 000 bu. Labor: 8 hr. /room/season Taxes and insurance.(2%): $28 per 20, 000 unit 192 Dry lime Initial cost (average) based on local labor: 1. 5 cents/bu. based on hired labor: 3 cents/bu. Years in service: 12 Salvage value: 10% Operational costs Dry lime cost: 1 cent/ bu. , based on price without resale or regeneration Labor: 12 hr. per room per season Taxes and insurance: (2%) Monoethanolamine (MEA) Approximate initial cost: 1/ 3 more expensive than water Scrubber: $1865 for 20, 000 bu. Years in service: 10 Salvage value: 10% Operational costs Power 1 cent/ kwh: $217 for 20, 000 bu. 3 cents/kwh: $650 for 20, 000 bu. Labor: 8 hr. /room/ season Taxes and insurance: (2%) 193 NaOH Initial cost: $1650 for 20, 000 bu. Years in service: 7 Salvage value: 10% Operational costs NaOH cost: $1000 for 20, 000 bu. Power 1 cent/ kwh: $25. 24 for 20, 000 bu. 3 cents/kwh: $75. 72 for 20, 000 bu. Labor: 1/2 hr. /day/room, $150 for 20, 000 bu. Taxes and insurance: (2%) APPENDIX 2 - 3 Oxygen Reduction Costs Eaves burner for 6000 bu. room Initial cost: $250 Years in service: 10 Salvage value: 10% Operational costs Fuel cost for 4 pulldowns based on 8. 4 cents/ lb. propane: $100 based on 2.83 cents/ lb. propane: $ 33. 7 Labor, 8 hr. /season: $1. 5/hr. Taxes and insurance: (2%) Tectrolgenerator1 as burner for 20,000 bu. Initial cost: $5000 Installation, 3 cents/bu.: $600 Years in service: 10 Salvage value: 10% 1The Tectrol unit includes the carbon dioxide scrubber that can be used, for the same basic cost for the pulldown period. 194 195 Operational costs for 4 pulldowns Power 1 cent/ kwh: $4 3 cents/ kwh: $12 Water: $ 5. 6 Fuel 2.83 cents/lb. propane: $ 32 8. 4 cents/ lb. propane: $ 90. 2 Labor: 8 hr. /season/ room Taxes and insurance (2%): $100 Arcat for 60, 000 bu. Initial cost: $ 3350 Shipment to Michigan $ 100 Installation $ 200 Years in service: 10 Salvage value: 10% Operational costs for 4pulldowns Power 1 cent/ kwh: $7. 2 3 cents/kwh: $21. 6 Water: $1.0 Fuel 2. 83 cents/lb. propane: $78. 5 8. 4 cents/ lb. propane: $ 233.0 Labor: 8 hr. /season/ room Taxes and insurance (2%): $67 196 Liquid N; Based on projected estimated costs for 10, 000 bu. room (Smock, 1965). The adjusted price for flushing with the liquid N; and flushing and cooling combination is based on comparison of the liquid N; operation costs with the estimated mechanical refrigeration pre- cooling (from 65 to 35° F) cost of 4. 18 cents per bu. 197 APPENDIX 2 -4 Complete Systems Annual Costs Tectrol - annual cost based on lease and operation costs The cost information provided by the Tectrol Division of Whirlpool Corporation is of a general nature and should be treated as such. It is based on 6 month operation, 2. 5 cu. ft. storage volume per bushel, electrical power at 1 cent or 3 cents per kwh, propane gas at 2. 83 cents/ lb (12 cents/ gallon), and four rooms with 15, 000 bu. in each room. The Tectrol service charge is 15 cents/ bu. /6 months. There is a first year installation and room leak test charge of approximately 3 cents/bu. The Tectrol service includes: six months service, four pulldowns, warantee, parts and on-call service, technical assistance and annual leak test. Costs for 60, 000 bu. CA room Installation: $ 1800 Operational costs Service charge: $9000 Power 1 cent/ kwh: $ 300 3 cents/kwh: $900 Water: $427 198 Fuel: 12 cents/gallon propane, $2400 Labor: 4 down periods a season, 8 hr. total Taxes and insurance: none The cost was also calculated assuming 40 cents/bu. saving on con- struction costs. For example, for 60, 000 the saving is $24, 000. The actual annual saving is $24, 000 X 0.05 = $1200 based on an- nual interest of 5%. Tectrol - annual cost based on purchase and operation costs The annual cost was based on a $5000 possible initial cost of the Tectrol generator for 20, 000 bu. (Pflug 1965) and the installation and operational costs provided by Tectrol Division of Whirlpool Corporation. Costs for 60, 000 bu. room Initial cost $15, 000 Installation costs, 3 cents/bu. $ 1, 800 Years in service: 10 Salvage value: 10% Operational costs Power 1 cent/ kwh $ 300 3 cents/kwh $ 900 Water $ 427 Fuel, 12 cents/gallon propane Labor, 4 down periods a season Taxes and insurance Arcagen for 60, 000 bu. Initial cost Shipment to Michigan Installation Years in service: 10 Salvage value: 10% Operational cost, based on 4pulldowns Power, Arcat 1 cent/ kwh: 199 3 cents/ kwh : Power, Arcosorb 1 cent/ kwh: 3 cents/kwh: Water Arcat Arcosorb Fuel, 12 cents/gallon propane Labor, 20 hr. per season Taxes and insurance (2%) $ 2, 400 $ 36 $ 300 $11, 900 $ 150 $ 500 $7.2 $21.6 $208 $624 $1.04 $13.5 $78.5 $35 $238 Eaves " Oxygen controller" for 6, 000 bu. Initial cost: $ 250 200 Years in service: 10 Salvage value: 10% Operational costs based on 4 pulldowns Fuel costs 8. 4 cents/ lb. propane: $100 2.83 cents/lb. propane: $ 33.7 Dry lime cost--because of extra load on scrubber 1. 5 cents/bu. was assumed: $90 Labor: 8 hr. /room Taxes and insurance (2%): $10 Based on projected estimated costs for 10, 000 bu. room (Smock 1965). The adjusted price for flushing is based on 4. 18 cents per bu. annual cost precooling 30° F with mechanical refrigeration. LITERATURE CITED Angelini, P. 1956. A study of CO; absorption equipment for CA fruit storages. M. S. Thesis, Michigan State University, East Lansing, Michigan (unpublished). Anon. 1959. Smith complete another cold store for Birds eye. Insula- tion Review (London) Feb. p. 24. Astarita, G. G. Marrucci and F. Gioia. 1964. The influence of carbon- ation ratio and total amino concentration on CO; absorption in aqueous MEA solutions. Chem. Eng. Sci. 12, 95. Awbery, J. H. 1927. The flow of heat in a body generating heat. Phil. Mag. (7), fl, 629. Bass, R. 1964. Defrosting methods used on blower evaporators. Proc. Food Eng. Conf. p. 67, April 1964, M.S.U. Benson, H. E., J. H. Field and W. P. Haynes. 1956. Improved process for CO; absorption uses hot carbonate solutions. Chem. Eng. Progr. 52, 433. Bird, R. B., W. E. Stewart and E. N. Lightfoot. 1965. Transport phenomena. 5th printing. John Wiley& Sons, New York. Blum, H. A., L. F. Stutzman, and W. S. Dodds. 1952. Absorption of carbon dioxide by sodium and potassium hydroxide. Ind. Eng. Chem. 3115, 2969. BRAB. Building Research Advisory Board. 1963. Cold storage fa- cilities; A guide to design and construction. National Academy of Sciences-National Research Council No. 1098. Buck, B. O., and A. R. S. Leitch. 1958. Carbon dioxide removal from natural gas. Petrol. Refiner 31, 241. Clarke, J. K. A. 1964. Kinetics of absorption of CO; in MEA solutions at short contact-time. Ind. Eng. Chem. Fundamental _3_, (3), 239. Clarke, R. G. 1964. Gels and methods of making. US. Patent 3, 191, 729. 201 202 Comstock, C. S. and B. F. Dodge. 1937. Rate of carbon dioxide absorp- tion by carbonate solutions in a packed tower. Ind. Eng. Chem. 22, 520. Cooper, C. M. 1945. Gas-scrubbing apparatus, U. S. Patent 2, 398, 345. Cooper, C. M., R. J. Christl and L. C. Peery. 1941. Packed tower per- formance at high liquor rates. Trans. Am. Inst. Chem. Eng. _3_7_, 979. Cryder, D. S. and J. O. Maloney. 1941. The rate of absorption of carbon dioxide in diethanolamine solutions. Trans. Am. Inst. Chem. Eng. 37, 827. Deed, D. W., P. W. Schutz, and T. B. Drew. 1947. Comparison of rectification and desorption in packed columns. Ind. Eng. Chem. _3_9_, 766. Dewey, D. H. and I. J. Pflug. 1965. Suggestions for the management and Operationw‘of your CA storage. Dept. of Horticulture and Food Science, Michigan State Univ., E. Lansing, Mich. Dow Chemical Company. 1965. Dow insulating panels for controlled at- mosphere apple storage. Plastics Dept. , Midland, Mich. Eaves, C. A. 1964. The effect of the chemical and physical properties of lime on scrubbing efficiency. Univ. of Mass. Amherst, Mass. Extension publication 422, 16. 1965. Kentville Nova Scotia. Personal correspondence. Eaves, C. A. and C. L. Lockhart. 1964. Oxygen controller for CA storage. Ann. report--Nova Scotia fruit growers' Assoc. Eckert, J. S., E. H. Foote, and R. L. Huntington. 1958. Pall rings-- new type of tower packing. Chem. Eng. Progr. 53, 70. Ergun, S. 1952. Fluid flow through packed columns. Chem. Eng. Prog. 48, 89. Fidler, J. C. 1961a. Physiological activity of volatile organic compounds produced by fruits. Intern. Inst. Refrig. Annex 1961-63, p. 379. 1961b. The measurement of carbon dioxide and oxygen con- centration. Inter. Inst. Refrig. Comm. 4, 241. 203 Foxboro Company, Foxboro, Mass. Feb. 1958. Wet and dry bulb hu- midity instrument-"wick types. Instruction sheet 3-110. Foxboro, Mass. May 1964. Dew point-recorders and control- lers. Bull. 11-11D. Fujita, S. and T. Hayakawa. 1956. Liquid-film mass transfer coefficients in packed towers and rod-like irrigation towers. Chem. Eng. (Japan) 20, 113. Furnas, C. C. and F. Bellinger. 1938. Operating characteristics of packed columns. Trans. Am. Inst. Chem. Eng. _3__4_, 251. Geleynse, J., W. P. F. R. I. Stellenbosch. 1962. Relative humidity- measurement and control in coldstores. The Deciduous Fruit Grower 1_2_, (5). General Electric Research Laboratory, Schenectady, N.Y. 1964. Arti- ficial gills are not for fish. Chem. Eng. ll (24), 94. Gray, H. E. and R. M. Smock. 1955. Farm refrigerated apple storages. Cornell Univ., Agr. Exp. Sta. Bull. 786. Greenwood, K. and M. Pearce. 1953. Removal of CO; from atmosphere air by scrubbing with caustic soda in packed towers. Trans. Inst. Chem. Eng. (London) _3_1_, 201. Gregory, L. B. and W. G. Scharmann. 1937. Carbon dioxide scrubbing by amine solutions. Ind. Eng. Chem. §_9, 514. Gregory, S. A., P. B. Herbert, and J. J. Lane. 1963. The removal of carbon dioxide by molecular sieve absorbent. Symposium on less common means of separation. (Inst. Chem. Eng. ). Guillou, R., H. B. Richardson. 1964. Humidity control in fruit cooling and storage. Univ. of Calif. Davis Agr. Ext. Serv. Nov. 1964. Hall, J. E. 1957. New scrubber for gas stores. Mod. Refrig. _€_>_0_, 322. Harte, C. R., E. M. Baker, and H. H. Purcell. 1933. Absorption of carbon dioxide in sodium carbonate-bicarbonate solutions. Ind. Eng. Chem., _2_5_, 528. Hersh, C. K. 1961. Molecular sieves. New York, Reinhold Publications Inc. 204 Honeywell Regulator Comp. Minneapolis, Minnesota. 1962. Dew Probe sensors. Specification data 90-1716. Hoogendoorn, J. C. 1963. A comparison of some carbon dioxide removal processes. Trans. Inst. Chem. Eng. 41, 264. Hukill, W. V. , and E. Smith. 1946. Cold storage for apples and pears. U.S. Dept. Agr. Circ. 740, February. Hygrodynamics, Inc., Silver Spring, Maryland. Humidity instrumentation and systems. Catalog SF-lB. Hunter, D. L. 1965. A doctor for the apple. The analyzer, Scientific and Process Instruments, Beckman Instruments, Inc. April. Hutcheon, N. B. 1958. Vapor problems in thermal insulation. ASHAE J. August. Jeffreys, G. V. , and A. F. Bull. 1964. The effect of glycine additive on the rate of absorption of CO; in sodium carbonate solutions. Trans. Inst. Chem. Engrs. g, 118. Jensen, W. F. 1965. New principles improving the CA system. Produce Marketing, April. Johansson, J. 1962. Concentration of volatiles in CA storage and their relation to some storage operation. Proc. Amer. Soc. Hort. Sci. 80, 137. 1963. Use of molecular sieve as a CO; adsorbent in experi- mental CA chambers. Am. Soc. Hort. Sci. 8_3, 172. Kane, J. J. 1961. Multi- stop delivery of frozen foods using liquid nitrogen refrigeration. Linde, Cryogenic Products Development Department (Unpublished) . Kendenburg, F. V. Jr. 1959. CA refrigerated storages a boon to fruit warehouses. Ind. Refrig., 137, (2), 22. Koch, H. A., L. F. Stutzman, H. A. Blum, and L. E. Hutchings. 1949. Gas absorption-liquid transfer coefficients for the carbon dioxide- air-water system. Chem. Eng. Progress, _4_5_, 677. Kohl, A. L. 1956. Plate efficiency with chemical reaction-absorption of carbon dioxide in Monoethanolamine solutions. Am. Inst. Chem. Eng. J. _2, 264. 205 Kondo, R. , and K. Fukuba. 1958. Absorption of CO; by MEA solutions in packed columns. Chem. Eng. (Japan) _2_2, 610. Kopelman, J., J. L. Blaisdell and I. J. Pflug. 1964. The influence of fruit size and collant velocity on the cooling of Jonathan apples in water and air. To be submitted at ASHRAE semi annual meeting, Houston, Texas January 1966. Lannert, J. W. 1964. Tectrol atmospheric control. Agr. Eng. 45, 318. Lentz, C. P. 1955. Humidification of cold storages: The Jacket System. Can. J. Technol. _3_3, 265. 1959. The Jacketed system for cold rooms. Proc. 10th Intern. Congr. Refrig. 3, 298. Lentz, C. P. , and W. H. Cook. 1955. Comparison of two types of high humitidy cold storage systems. Proc. 9th Intern. Congr. Refrig. , 2, p. 5.039. Lentz, C. P., and U. Nakano. 1961. Predicting air friction pressure loss in shallow ducts. ASHRAE J. _3_, February. p. 32. Lentz, C. P., and C. W. R. Philips. 1959. Use of the jacketed room system for fresh fruit and vegetable storage. Proc. 10th Intern. Congr. Refrig. p. 309. Leva, M. 1953. Tower packing and packed tower design, 2nd ed. U. S. Stoneware Co. , Akron, Ohio. 1955. Gas absorption in beds of rings and saddles. Am. Inst. Chem. Eng. J. l, 224. Lorenzen, R. T. 1963. A balancing technique for moisture control in structural components of CA storages. Winter meeting Am. Soc. Agr. Eng. Paper No. 63-905. December, Chicago. Lotz, W. A. 1964. Vapor barrier design for cold storage application. ASHRAEJ., é, Dec. p. 46. Mann, G. 1954. The control of relative humidity in cold stores. 1. The loss of water from perishable produce during storage. The Inst. Refrig. , London. 1953-54 Session March 16. 1955. Control of loss of weight by design of the installation. Proc. 11th Intern. Congr. Refrig. Comm. 4. 206 1958. The use of ethanolamine in scrubbers for fruit stores. Mod. Refrig. 61, 990. 1960. Control of conditions in fruit and vegetables stores. J. Inst. Agr. Eng. _1_6, October. Mayland, B. J. 1964. Removal of carbon dioxide from gaseous mixtures. U. S. Patent 3, 144, 301. Mayland, H. B. 1956. Industrial engineering handbook. p. 7-84. McGraw- Hill Book Co. McGarvey, A. R. 1964. Wrap your cold storage room in a vapor barrier. ASHRAE J. 6, Dec. p. 42. Meigh, D. F. 1956. Volatile compounds produced by apples. 1. Alde- hydes and Ketones. J. Sci. Food Agr. 1, 396. 1957. Volatile compounds produced by apples. 2. Alcohols and esters. J. Sci. Food Agr. 8, 313. Meyer, A. 1961. Sulzer CO; washing plant for the cold storage of fresh fruit. Intern. Inst. Refrig. Comm. 4, p. 267. Meyer, C. S. 1964. Inside out design developed for low temperature building. ASHRAE J. g, April, p. 50. Nelson, K. E., R. Guillon. 1960. Highly marketable grapes maintained by controlled humidity. ASHRAE J. 2, 57. Palmer, R. 1959. Special report to CA storage operators. Am. Fruit Grower, 19, (8), 20. Perry, J. H 1963. Chemical engineering handbook. 4th ed. McGraw- Hill, New York. Phillips, W. R., C. P. Lentz, E. A. Rooke and W. M. Rutherford. 1961. The use of jacketed room system for the storage of apples. Canadian Refrig. Air Conditioning) Dec. Pflug, I. J. 1956. Breather bag for CA fruit vegetables storages. Agri- cultural Engr. _31, (7), 476. 1960. Oxygen reduction in CA storages: A comparison of water vs caustic soda absorbers. Mich. State Univ. Agr. Exp. Sta. Quart. Bull. 33, 455. 207 1961. Design of water type carbon dioxide absorption system for fruit storage. ASHRAE J. _3, 65. 1965. Mich. State Univ. Personal Communication. Pflug, I. J., and J. L. Blaisdell. 1963. Method of analysis of precooling data. ASHRAE J. _5_, November, p. 33. Pflug, I. J., and D. H. Dewey. 1955. Controlled atmosphere storage. Agrigultural Engr. 35, (3), 171. 1956. A theoretical relationship of the variables affecting the operation of CA storage rooms. Mich. State Univ. Agr. Exp. Sta. Quart. Bull. 39, 353. 1959. Construction for controlled atmosphere apple storage. Agr. Eng. _4_0_, (2), 80. Pflug, I. J., and F. W. Southwick. 1954. Air leakage in controlled at- mosphere storage. Agr. Eng. §_5_, (9), 635. Pflug, I. J., P. Angelini, and D. H. Dewey. 1957a. Fundamentals of CO; absorption as they apply to CA storages. Mich. State Univ. Agr. Exp. Sta. Quart. Bull. fl, 131. Pflug, I. J., M. W. Brandt and D. H. Dewey. 1957b. An experimental tilt up concrete building for the controlled-atmosphere storage of apples. Mich. State Univ. Agr. Exp. Sta. Quart. Bull. _32, 505. Pflug, I. J., D. H. Dewey and M. W. Brandt. 1957c. An absorber for the removal of carbon dioxide from CA storage. Mich. Agr. Expt. Sta. Quart. Bull. 40, 131. Pflug, I. J., J. L. Blaisdell and J. Kopelman. 1965. Developing tem- perature-time curves for objects that can be approximated by a sphere, infinite plate, or infinite cylinder. ASHRAE Transactions 31. Poms, A. I. 1965. Controlled atmosphere enclosures. Koppers Company. Panel dept. Detroit, Mich. February. Richards, G. M., G. A. Ratcliff and P. V. Danckwerts. 1964. Kinetics of CO; absorption-3. First order reaction in a packed column. Chem. Eng. Sci. _12, 325. Robb, W. L. , Manager, Chemical Process Studies. General Electric Com- pany. 1964. Personal correspondence. 208 Sainsbury, G. F. 1961. Cooling apples and pears in storage rooms. U. S. Dept. Agr. AMS 474, September. 1962. Cooling apples in pallet boxes. U. S. Dept. Agr. AMS 532, August. Sherwood, T. K. and F. A. L. Holloway. 1940. Performance of packed towers-liquid film data for several packing. Trans. Am. Inst. Chem. Eng. _3_6, 39. Sherwood, T. K. and Pigford, R. L. 1952. Absorption and extraction. McGraw-Hill, Inc. New York, N.Y. Shulman, H. L. and J. J. DeGauff, Jr. 1952. Mass transfer coefficients and interfacial areas for l-inch Rashing rings. Ind. Eng. Chem. 3;, 1915. Shuman, E. C. 1961. Private communication to BRAB cold storage com- mittee 1961. Smock, R. M. 1960. Water scrubbing in CA rooms. Cornell Univ. Dept. of Pomology, Mem. 5-508. 1961. Methods of scald control on the apple. Cornell Expt. Sta. Bull. 970, October. 1962. Gas sealing the CA room. Cornell Univ. Dept. of Pomology, Mem. S-528. 19 6 5. P ersonal communication. Smock, P. M. and G. D. Blanpied. 1960. Handbook for CA rooms. Cornell Univ. Dept. of Pomology. Mem. S-504. Smock, R. M. and G. D. Blanpied. 1964. Lime scrubbing, water scrub- bing and externally generated atmospheres compared in CA storage. Univ. of Mass. Amherst, Mass. Extension publication 422, p. 22. 1965. The use of nitrogen for cooling, flushing, and maintain- ing atmosphere in CA rooms. Proc. 1965 Annual Meeting, New York State Hort. Soc. Smock, R. M. and VanDoren, A. 1941. CA storage of apples. Cornell University Agr. Exp. Sta. Bull. 762. 209 Southwick, F. W. and J. W. Zahradnik. 1958. CA apple storage. Univ. of Mass. Amherst, Mass. Extension Bull. 322. Spector, N. A. and B. F. Dodge. 1946. Removal of carbon dioxide from at- mospheric air. Trans. Am. Inst. Chem. Eng. _42, 827. Strutz, W. 1964. Evaporator coils and humidity in fruit and vegetable storages. Proc. Food Eng. Conf. p. 53. April 1964. M. S. U. Teller, A. J. and R. E. Ford. 1958. Packed column performance of carbon dioxide-monoethanolamine system. Ind. Eng. Chem. 52, 1201. Tepe, J. B. and B. F. Dodge. 1943. Absorption of carbon dioxide by sodium hydroxide solutions in a packed column. Trans. Am. Inst. Chem. Eng _3_9, 255. Thomas, O. O. Atlantic Research Corporation. 1965. Personal correspond- ence. Turk, A. 1955. Air purification in refrigerated storage in the U. S. A re- view. Proc. 9th Intern. Congr. Refrig. Paris, p. 41. U. S. Dept. Agr. 1954. The commercial storage of fruits, vegetables, and florist nursery stocks. Agr. Handbook, 66. Walter, J. F. and T. K. Sherwood. 1941. Gas absorption in bubble-cap columns. Ind. Eng. Chem. 33, 493. Wolf, J. and J. Kuprianoff. 1961. Apparatus for control of storage and simultaneous measurement of the gas exchange of the stored produce. Intern. Inst. Refrig. Comm. 4, p. 34. Wright, N. 1961. A program controlled environmental plant growth chamber. Univ. Ariz. Agr. Exp. Sta. Tech. Bull. 148. Yoshida, F. , and T. Koyanagi. 1958. Liquid phase mass transfer rates and effective interfacial area in packed absorption columns. Ind. Eng. Chem. 5_()_, 365. Zahrandnik, J. W. 1964. Functional specifications for CA instrumentation controlled atmosphere storage. Univ. of Mass. Amherst, Mass. Ex- tension publication 422, p. 5. 210 Zahradnik, J. Whand F. W. Southwick. 1958. Design details and per- formance characteristics of a Douglas fir plywood apple storage. Univ. of Mass. Amherst, Mass. Exp. Sta. Bull. 505. Zahradnik, J. W., F. W. Southwick, and J. M. Fore. 1958. Internal atmosphere movement in CA apple storage. Agr. Eng. 32, (5), 288.