- Email: [email protected]
Modification and testing of a liquid nitrogen refrigerated container for the distribution of fresh red meat
Food Research International 33 (2000) 759±765
www.elsevier.com/locate/foodres
Modi®cation and testing of a liquid nitrogen refrigerated container for the distribution of fresh red meat M.N.N. Habok a, D.S. Jayas a,*, R.A. Holley b a
Department of Biosystems Engineering, University of Manitoba, Winnipeg, MB, Canada R3T 5V6 b Department of Food Science, University of Manitoba, Winnipeg, MB, Canada R3T 2N2 Received 23 September 1999; accepted 24 March 2000
Abstract A previously-designed insulated container was modi®ed to improve the hold period (variable) nitrogen (N2) use of the cryogenic system used to cool the container. Although fresh meat storage life is longest at ÿ1.50.5 C, the target temperature range for the system was changed to ÿ1.00.5 C to eliminate any economic loss caused by the freezing of meat. A centrifugal fan, with its motor mounted outside the container, provided a uniform temperature within the container and lowered the hold period N2 use of the system. Chilling of the product from an initial temperature of 7.0 C to ÿ1.0 C required on average 33.7, 31.7, 34.7, and 30.6 kg of N2 at outside temperatures of 30, 15, 0, and ÿ15 C, respectively. Average hold period N2 use for maintaining the required temperature was 4.8, 2.8, 1.2, and 0.1 kg/h at outside temperatures of 30, 15, 0, and ÿ15 C, respectively, and these were an improvement over the previous design. The overall maximum temperature of the container during the trials was ÿ0.3 C, while the overall minimum was ÿ1.4 C. A vacuum insulated transfer line decreased hold period N2 use by 51% in comparison with a steel braided transfer line. A simulated N2 failure of the system resulted in an average temperature increase of 0.5oC/h within the container. # 2000 Elsevier Science Ltd. All rights reserved.
1. Introduction Fresh red meat is more pleasing to consumers than frozen meat as the meat tissues do not suer detrimental eects due to ice crystal formation (Taylor, 1985). Demand for red meat is increasing due to the growth of exports to countries in the Asia-Paci®c region and Mexico (Anon., 1995). Shay and Egan (1990) reported that the preparation, sea transportation, and distribution of meats to these distant markets can take up to 10 weeks. In the current system, animals are slaughtered and then subsequently formed into primal and sub-primal cuts (15±20 kg) which, upon arrival at the retail outlet, require further processing to create retail-ready portions (Bailey, 1997). A small amount of meat is prepared retail-ready at the processing plants and distributed to local markets because fresh meat is a perishable product.
* Corresponding author. Tel.: +1-204-474-9868; fax: +1-204-4747525. E-mail address: [email protected] (D.S. Jayas).
An extension of shelf life is important to increase the distribution radius of fresh red meat. The shelf life attainable using conventional vacuum packaging cannot dependably achieve the shelf life needed for export or for distribution to distant markets. Controlled atmosphere packaging using saturating volumes of 100% CO2 with strict temperature control at ÿ1.50.5 C could enable worldwide distribution of fresh meat (Gill & Phillips, 1993; Holley, Delaquis, Gagnon, Doyon & GarieÂpy, 1993). A subset of this technology called master packaging, where retail-ready portions are prepared and wrapped with O2-permeable ®lms and then grouped together and overwrapped in gas impermeable packages, may ®nd application in local and intra-continental fresh meat distribution where labour and space for ®nal product preparation at retail are not available economically (Tewari, Jayas & Holley, 1999). Master packaging successfully combines the bacteriostatic eect created by modi®ed atmosphere packing with CO2 and provides for the oxygenation of the muscle pigment myoglobin (producing the consumer-desired red colour) through the utilization of double-layer packaging. However, a system which can successfully and economically
0963-9969/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved. PII: S0963-9969(00)00079-X
760
M.N.N. Habok et al. / Food Research International 33 (2000) 759±765
maintain the temperature requirements and work in conjunction with master packaging technology has yet to be developed. The objectives of this study were: (i) to modify a container designed by Bailey, Jayas, Holley, Jeremiah and Gill (1997) which uses liquid N2 as the refrigerating medium, to reduce the N2 use of the system, (ii) to test the modi®ed container's ability to chill and maintain meat temperatures over a range of operating conditions, and (iii) to compare the use of a steel braided transfer line for N2 delivery to the container with a vacuum insulated transfer line. 2. Materials and methods 2.1. Design modi®cations and instrumentation 2.1.1. Experimental apparatus A container previously designed and tested by Bailey et al. (1997) was used in this study. The container (Model C-54, Les Contenants Xactics LteÂe, Joliette, PQ) was equipped with a stainless-steel shelving-system, capable of holding 36 master trays at nine levels, with four master trays at each level (Fig. 1). The shelving unit is able to accommodate master trays with dimensions of 50838160 mm, which contain packaged red meat. The liquid N2 refrigerating medium used to chill the container was supplied from a pressurized tank (Dura Series, MVE, New Prague, MN) via a transfer line. Liquid N2 was dispersed in the head space with the use of copper piping and agricultural sprayer nozzles. The design by Bailey et al. relied on the use of six fans loca-
ted in the headspace at the top of the container to circulate the cooling medium throughout the container. Regulation and control of the N2 injections was accomplished through a computer control algorithm (Bailey, 1997) which delivered N2 to the container via a solenoid valve (Model 8263G206 LT, ASCOlectric Ltd., Brantford, ON) located in the transfer line. The container was placed within an environmental chamber (Model C1010, Conviron, Controlled Environments Ltd., Winnipeg, MB) to enable testing of the container over a range of outside temperatures. 2.1.2. Preliminary design modi®cations The heat generated by the fans in the former design contributed to the hold period N2 use of the system. Therefore, this study included an investigation into circulating N2 within the container without the use of fans, using dierent nozzle positions and dierent nozzle types. However, all of the designs tested created temperature gradients within the container and the temperature values were outside the desired range. 2.1.3. Final design of the container A fan was added to enhance uniformity of the temperatures within the container and to improve the accuracy of the temperatures in relation to the target temperature of ÿ1.50.5 C. Bailey et al. (1997) determined that hold period N2 use increased as the number of fans increased, due to heat generated by the fan motors situated inside the container. Therefore, it was decided that the addition of a fan would only prove bene®cial if the additional heat generated by its motor did not enter the container. The resulting fan design
Fig. 1. Schematic of the container used by Bailey (1997) but showing the installation of copper piping and nozzle placement as modi®ed in the present study.
M.N.N. Habok et al. / Food Research International 33 (2000) 759±765
(Fig. 2) was the ®nal fan design for the present system. It consisted of a centrifugal fan encased within a metal enclosure, drawing N2 in from an open bottom. The metal enclosure discharged N2 from the fan into a plenum, consisting of a wooden box, which, in turn, discharged air into two circular (each with a diameter of 100 mm) metal conduits extending vertically downwards from the plenum. The circular conduits were connected to the central conduit within the shelving structure of the container. The central conduit was perforated with a series of 19 mm diameter holes to allow the N2 gas to ¯ow outward from the central conduit over the meat and up along the side conduits (Fig. 2). Nitrogen was delivered to the nozzles within the container with 9.5 mm diameter copper piping and was injected using four V-pattern nozzles (Model 11006VS, Even-Spray and Chemicals Ltd., Winnipeg, MB) which pointed upwards at the level of second layer from the bottom shelf (Fig. 1). 2.1.4. Instrumentation Instrumentation was needed to monitor the temperatures and N2 use of the container. All data were written to the hard drive of a 486 PC by a Quick BASIC program. To measure N2 use, the N2 tank was placed on a ¯oor scale (Model:2136, Mettler-Toledo Inc., Burlington, ON) and its mass was taken every minute by the QuickBASIC control algorithm. An RS 232 connection was used to transfer the mass data to the hard disk. Forty-four thermistors (Model: 44034, Omega Engineering Inc., Stamford, CT) (each with an accuracy of0.1 C) were used to measure the meat temperatures throughout the container and two thermistors were used to measure the temperature of the Conviron chamber. Two thermistors, located inside the container on the second from top and bottom shelves, were used as sensors for the control algorithm. A data acquisition/ control unit (Model 3852A, Hewlett-Packard Co., Loveland, CO) was used to measure the resistance of
761
the thermistors and in turn, relayed the information to the computer through a serial port. Activation of the solenoid valve was controlled by the control algorithm. 2.1.5. Modi®cations to the temperature control algorithm The control algorithm used by Bailey (1997) relied on the concept that the temperature of the product would converge to the cyclic mean temperature of the surrounding air inside the container. However, the meat has to be chilled to the target temperature within approximately the ®rst 2 h of storage to ensure the microbiological safety of the meat. Preliminary testing of the container, following the design changes in this study using pork roasts and simulated master trays, showed that it took approximately 20 h to reach the target temperature. This was because the algorithm would prematurely shut o the solenoid before the meat attained the target temperature. Therefore, the temperature control algorithm was modi®ed to use actual meat temperatures to control the activation of the solenoid in the chilling portion during storage to bring the meat to the desired temperature in the shortest possible time. In addition, the target temperature for the meat was raised to ÿ1.00.5 C to ensure that none of the meat would freeze. During our preliminary testing, some surface ``crusting'' occasionally occurred at ÿ1.50.5 C. Therefore, it was decided that an additional margin of safety should be built into the target temperature of the control algorithm. The formerly used control algorithm was designed so that N2 would only be injected throughout the last 40 s of the 60 s time cycle to allow for an equilisation of temperatures within the container and to allow the instrumentation system to take temperature measurements in the ®rst 20 s of the cycle. However, the data acquisition unit used in this study (Model 3852A, Hewlett-Packard, Avondale, PA) does not require the additional time to take readings, therefore temperature
Fig. 2. Schematic of the ®nal fan design used in this study and the ¯ow of N2 shown by arrows inside the container.
762
M.N.N. Habok et al. / Food Research International 33 (2000) 759±765
readings can be taken simultaneously with N2 injections. Furthermore, if the activation of the solenoid relies solely upon the meat temperature, then an equilisation period is not needed. To ensure that none of the meat would freeze during the chilling period, the minimum meat temperature was not allowed to drop below ÿ1.75 C. In addition, the constraint preventing the air temperature within the container to fall below ÿ20.0 C ensured that the meat would not freeze due to heat transfer from the air to the meat. This value for the minimum air temperature was decided by trial and error. 2.1.6. Simulated master trays Testing of the ®nal container design was conducted with meat samples (three 2 kg pork butt roasts) contained within each simulated master tray. The design of simulated master trays was based on the geometry of an actual master tray so that the ¯ow characteristics of N2 within the container and the heat transfer within the trays could be modelled in a realistic manner. The design of the simulated master tray (Fig. 3) consisted of an aluminum roasting pan (47.336.56.7 cm), with an overlying wire frame, bent to create a head space of air above the 6 kg meat per tray. The tray and frame were enclosed in an unsealed plastic bag (5687 cm) to allow positioning of a thermistor for meat temperature measurement. The head space between the meat and the plastic bag simulated the dead space created within a master package by the back-¯ushed atmosphere. This dead air space aected the heat transport characteristics of the tray in a manner similar to an actual master tray. The thermistor used to measure the meat temperature was passed through the plastic wrap on the meat and was attached to the upper meat surface without penetrating it and was held in place with duct tape. The surface temperature of the meat was chosen as it is the most sensitive indicator of muscle tissue freezing. 2.2. Testing procedures 2.2.1. Temperature and nitrogen use tests The Conviron chamber was used to chill the container and the meat samples to 71.5 C because this was
Fig. 3. Schematic of a simulated master tray.
considered an average temperature of meat entering the chilling process at a processing facility (Gill & Phillips, 1993). This diered from the testing procedure used by Bailey et al. (1997) in which saline water bags, simulating meat samples, were at an initial temperature of 101.5 C. After the initial temperature had been maintained brie¯y (45 min), the container door was closed and the Conviron chamber was then set to the desired outside temperature for the test to be conducted. The liquid N2 release valve of the N2 tank was then opened and the temperature control algorithm was started. The preliminary tests of the container were conducted for 8 h (Bailey et al.), however, tests conducted with the ®nal design of the container lasted for approximately 2 days to ensure that the system had reached a stable temperature. Three replicates were completed for the outside temperatures of 30, 15, 0, and ÿ15 C. 2.2.2. Comparison of vacuum insulated and steel braided transfer lines Two replicates at a chamber temperature of 15 C were conducted to compare the eect on N2 use of using a vacuum insulated transfer line with an uninsulated steel braided transfer line. The latter was used because of the low cost of purchase, its durability, ¯exibility and ease of use during tank installation. Each of these replicates were evaluated for about 24 h. 2.2.3. Failure testing Failure testing was conducted to simulate a N2 failure and to determine the resulting eect on the container temperature. The container and meat samples were equilibrated to ÿ1.01.5 C using the Conviron chamber. After this temperature was reached, the container door was closed and the Conviron chamber was then set to 30 C. Temperatures throughout the container were then recorded for 8 h and three replicates were conducted. 2.2.4. Method of data analysis Temperature history and N2 use versus time were studied using observations at 60 s intervals. The means of maximum, average, and minimum temperatures of meat samples were determined from 5.5 h after initial chilling from 7 C until the trial was completed (Table 1), because the meat samples had reached temperature equilibrium at that point. Nitrogen use in terms of its pre-cool (®xed) and hold (variable) components were determined. The former refers to the quantity of N2 required to chill the contents of the container from 7 C to the desired temperature, while the latter represents the amount of N2 needed to maintain the desired temperature. Pre-cool period N2 use was determined as the amount of N2 consumed up to the beginning of the horizontal section of the curve for N2 use versus time (Fig. 4). Hold period N2 use was
M.N.N. Habok et al. / Food Research International 33 (2000) 759±765
calculated by performing a linear regression on the N2 use data over time from 5.5 h until test completion. The horizontal portion of the curve was not used in calculating N2 consumption. After 5.5 h, the N2 use by the container had reached a constant value (Fig. 4). The slope of the line between N2 use versus time represented the hold period N2 use. Table 2 shows the experimental values for pre-cool and hold period N2 use. 3. Results and discussion 3.1. Temperature and nitrogen use data 3.1.1. Temperature There were no temperature gradients detectable within the container during testing. The temperature range measured in the container was similar for each outside temperature tested. However, it appeared that there was a slight rise in the temperature range as the outside temperature decreased, i.e., the maximum values shifted approximately 0.1 to 0.2 C higher at 0 and ÿ15 C than at the higher outside temperatures. The present design has the additional advantage of being able to operate at sub-zero outside temperatures without freezing of the meat samples because of heat generated by the fan. As the target temperature range for this study (ÿ1.00.5 C) and the previous study by Bailey et al. (1997), (ÿ1.50.5 C), are not the same, a direct comparison of temperature uniformity between the two cannot be made. However, the temperature Table 1 The maximum, average, and minimum temperatures of meat samples within the container from 5.5 to 48 ha Outside temperature Trial number ( C)
Temperature ( C)
Table 2 Experimental values of pre-cool and hold period nitrogen use
30
ÿ1.2 ÿ1.1 ÿ1.1
1 2 3
33.0 30.6 37.4
4.5 5.0 5.0
15
ÿ0.9 ÿ0.8 ÿ0.8
ÿ1.0 ÿ1.0 ÿ1.1
1 2 3
25.2 31.6 38.2
2.8 2.7 2.8
0
ÿ0.5 ÿ0.3 ÿ0.4
ÿ0.9 ÿ0.9 ÿ0.9
ÿ1.3 ÿ1.2 ÿ1.2
1 2 3
38.6 31.2 34.2
1.2 1.2 1.2
ÿ15
ÿ0.3 ÿ0.2
ÿ0.8 ÿ0.8
ÿ1.0 ÿ1.1
1 2 3
33.2 29.6 29.0
0.0 0.1 0.1
Steel braided transfer linea 15 1 2
41.0 38.4
5.5 5.5
1 2 3
ÿ0.6 ÿ0.7 ÿ0.5
ÿ0.9 ÿ0.9 ÿ0.9
ÿ1.4 ÿ1.1 ÿ1.1
15
1 2 3
ÿ0.7 ÿ0.5 ÿ0.4
ÿ1.0 ÿ0.9 ÿ0.9
0
1 2 3
ÿ0.4 ÿ0.4 ÿ0.3
1 2 3
a
Fig. 4. Typical N2 use data from a replicate trial at an outside temperature of 15 C. Pre-cool N2 consumption was calculated using data on the ordinate from 0 to the beginning of the horizontal portion of the curve indicated by an arrow.
Hold period nitrogen use (kg/h)
Minimum
Steel braided transfer lineb 15 1 2
3.1.2. Nitrogen use Table 2 shows the pre-cool period N2 use values for each test and these average values for trials at 30, 15, 0, and ÿ15 C were 33.7, 31.7, 34.7, and 30.6 kg, respectively. Pre-cool period N2 use was similar for all temperatures tested (ScheeÂ's mean comparison test; =0.05). This was expected because the amount of meat was nearly the same for all trials and heat transfer between the container and the outside air would be negligible during the 2 h period of initial chilling. Bailey et al. (1997) observed an average pre-cool period N2 use of 18.1 kg, while the data from this study
Pre-cool nitrogen use (kg)
Average
ÿ15a
range was smaller with the present design than with the previous design (Table 3). Also, the variability in temperature range with a change of outside temperature was more consistent with the present design (Table 3).
Trial number
Maximum
30
763
Trials conducted for 24 h with the same meat used for the three replicates. b Steel braided transfer line as opposed to the vacuum insulated transfer line used in all other tests.
Outside temperature ( C)
a Steel braided transfer line used as opposed to the vacuum insulated transfer line used in all other tests.
764
M.N.N. Habok et al. / Food Research International 33 (2000) 759±765
Table 3 A comparison of the average container temperature ranges for dierent outside temperatures between this study and the study by Bailey (1997) Outside temperature ( C)
Average temperature range ( C) Present study
Study by Bailey (1997)
30 15 0 ÿ15
0.6 0.6 0.7 0.8
1.2 1.0 0.6 0.4
yielded an average value of 32.7 kg. The increased precool period N2 use for the present design is attributable to the fact that Bailey et al. conducted trials using saline bags equivalent to 3.6 kg of meat per master tray, while in this study simulated master trays contained approximately 6 kg of meat per tray. The increased amount of meat within the container used for the present study increased the pre-cool period N2 requirement for initial chilling. Another factor that contributed to the dierence in pre-cool period N2 use was that the initial meat temperatures used by Bailey et al. were 3 C higher than those used here. Hold period N2 use is shown in Table 2 and as expected, the N2 use during the hold period increased as the outside temperature increased. This was due to the higher rate of heat conduction occurring through the transfer line and container walls at the higher outside temperatures. As outside temperature increased, hold period consumption of N2 increased in a proportional manner. Increasing the outside temperature from 15 to 30 C resulted in the N2 use during holding to almost double. Bailey et al. (1997) reported achieving average hold period N2 use values of 5.5, 4.0, 2.6, and 0.9 kg/h for 30, 15, 0, and ÿ15 C outside temperatures, respectively, using the four fan design. In this study, values of 4.8, 2.8, 1.2, and 0.1 kg/h for 30, 15, 0, and ÿ15 C outside temperatures, respectively, were determined for hold period N2 use. Therefore, it can be concluded that the present design represented an improvement in the eciency of hold period N2 use by comparison with the previous design, especially since the storage capacity was almost doubled in the newer design. 3.2. Comparison of vacuum insulated and steel braided transfer lines A comparison of N2 use and container temperatures resulting from installation of either a steel braided or vacuum insulated transfer line was made at an outside temperature of 15 C (Tables 1 and 2). Average pre-cool N2 use was similar for both types of transfer lines (LSD mean comparison test; =0.05) with values of 31.7 and
Table 4 Average rate of meat temperature increase after a nitrogen failure at an outside temperature of 30 C Trial number
1 2 3
Average rate of temperature increase ( C/h) Present study
Study by Bailey (1997)
0.4 0.5 0.5
1.9 2.0 2.0
39.7 kg observed for tests with vacuum insulated and steel braided transfer lines, respectively (Table 2). In contrast, hold period N2 use was signi®cantly aected by the type of transfer line used. Average values for hold period use were 2.8 and 5.5 kg/h, respectively, for the vacuum insulated and steel braided transfer lines. This result was expected because the vacuum insulated transfer line would reduce heat conduction. It was important to quantify changes in N2 use using dierent lines because the ¯exible (braided) line oers convenience and durability during hook up of the tank. Nonetheless, data showed that a vacuum insulated transfer line should be used. A comparison of the data from studies of the two types of transfer lines showed that although the minimum and average temperatures were similar, the tests with the steel braided transfer line had a maximum temperature which was approximately 0.3 C higher than that obtained for the vacuum insulated transfer line. It is unknown why this phenomenon occurred. 3.3. Nitrogen failure A linear regression was performed on the average temperatures of the meat samples during a test to determine the rate of temperature increase following a simulated N2 failure (Table 4). The present system had an average rate of temperature increase following a N2 failure at approximately 1/4 of the values determined by Bailey et al. (1997). 4. Conclusions The following conclusions can be drawn from this study: . Fans were essential to achieve the temperature uniformity required throughout the container (ÿ1.00.5 C). A single fan with the motor mounted outside the container and nozzles injecting N2 upwards along the side conduits at a level near the second-from-bottom shelf, served as the best design in terms of N2 use and temperature uniformity.
M.N.N. Habok et al. / Food Research International 33 (2000) 759±765
. The container operated inside the temperature range (ÿ1.00.5 C) for the average and minimum container temperatures, however, slightly outside the range as far as the maximum temperature is concerned. Maximum and minimum temperature readings for the container were ÿ0.3 and ÿ1.4 C, respectively. . Use of a vacuum insulated transfer line was bene®cial with regard to the N2 use. Use of a vacuum insulated transfer line reduced hold period N2 use almost in half compared to the braided line. . The present design signi®cantly improved the hold period N2 use over the former design by Bailey et al. (1997) and would serve as a more economically feasible method for transport of retail-ready fresh meat.
Acknowledgements The authors thank Messrs. Dale Bourns, Matt Macdonald, and Jack Putnam for their technical assistance and the Natural Science and Engineering Research Council of Canada for partial funding of this project. We thank Dr. C.O. Gill and Dr. L.E. Jeremiah for their technical advice during execution of this research project.
765
References Anon. (1995). Industry pro®le Ð fresh red meat. Report commissioned to Kline & Co. Inc. New York, by AlliedSignal Inc., Morristown, NJ (http://www.asresin.com/plastic/trailer.map). Bailey, C. G. (1997). Design, fabrication, and testing of a returnable, insulated, nitrogen-refrigerated, controlled-atmosphere, shipping container for distribution of fresh red meat. Unpublished MSc thesis, Department of Biosystems Engineering, University of Manitoba, Winnipeg, MB. Bailey, C. G., Jayas, D. S., Holley, R. A., Jeremiah, L. E., & Gill, C. O. (1997). Design, fabrication, and testing of a returnable, insulated, nitrogen-refrigerated shipping container for the distribution of fresh red meat under controlled CO2 atmosphere. Food Research International, 30(10), 743±753. Gill, C. O., & Phillips, D. M. (1993). Assessment of the preservative capabilities of storage and distribution processes. Technical Bulletin 1993-9E. Canada: Research Branch, Agriculture Canada. Holley, R. A., Delaquis, P., Gagnon, J., Doyon, G., & GarieÂpy, C. (1993). Modi®ed atmosphere packaging of fresh pork (Food Research and Development Centre Agriculture and Agri-Food Canada, St. Hyacinthe, PQ). Alimentech, 6(3), 16±17. Shay, B. J., & Egan, A. F. (1990). Extending retail storage life of beef and lamb by modi®ed atmosphere packaging. Food Australia, 42(8), 399±400, 404. Tewari, G., Jayas, D. S., & Holley, R. A. (1999). Centralized packaging of retail meat cuts: a review. Journal of Food Protection, 62(4), 418±425. Taylor, A. A. (1985). Packaging fresh meat. In R. Lawrie, Developments in meat science (pp. 89±113). New York: Elsevier Applied Science Publishers.
www.elsevier.com/locate/foodres
Modi®cation and testing of a liquid nitrogen refrigerated container for the distribution of fresh red meat M.N.N. Habok a, D.S. Jayas a,*, R.A. Holley b a
Department of Biosystems Engineering, University of Manitoba, Winnipeg, MB, Canada R3T 5V6 b Department of Food Science, University of Manitoba, Winnipeg, MB, Canada R3T 2N2 Received 23 September 1999; accepted 24 March 2000
Abstract A previously-designed insulated container was modi®ed to improve the hold period (variable) nitrogen (N2) use of the cryogenic system used to cool the container. Although fresh meat storage life is longest at ÿ1.50.5 C, the target temperature range for the system was changed to ÿ1.00.5 C to eliminate any economic loss caused by the freezing of meat. A centrifugal fan, with its motor mounted outside the container, provided a uniform temperature within the container and lowered the hold period N2 use of the system. Chilling of the product from an initial temperature of 7.0 C to ÿ1.0 C required on average 33.7, 31.7, 34.7, and 30.6 kg of N2 at outside temperatures of 30, 15, 0, and ÿ15 C, respectively. Average hold period N2 use for maintaining the required temperature was 4.8, 2.8, 1.2, and 0.1 kg/h at outside temperatures of 30, 15, 0, and ÿ15 C, respectively, and these were an improvement over the previous design. The overall maximum temperature of the container during the trials was ÿ0.3 C, while the overall minimum was ÿ1.4 C. A vacuum insulated transfer line decreased hold period N2 use by 51% in comparison with a steel braided transfer line. A simulated N2 failure of the system resulted in an average temperature increase of 0.5oC/h within the container. # 2000 Elsevier Science Ltd. All rights reserved.
1. Introduction Fresh red meat is more pleasing to consumers than frozen meat as the meat tissues do not suer detrimental eects due to ice crystal formation (Taylor, 1985). Demand for red meat is increasing due to the growth of exports to countries in the Asia-Paci®c region and Mexico (Anon., 1995). Shay and Egan (1990) reported that the preparation, sea transportation, and distribution of meats to these distant markets can take up to 10 weeks. In the current system, animals are slaughtered and then subsequently formed into primal and sub-primal cuts (15±20 kg) which, upon arrival at the retail outlet, require further processing to create retail-ready portions (Bailey, 1997). A small amount of meat is prepared retail-ready at the processing plants and distributed to local markets because fresh meat is a perishable product.
* Corresponding author. Tel.: +1-204-474-9868; fax: +1-204-4747525. E-mail address: [email protected] (D.S. Jayas).
An extension of shelf life is important to increase the distribution radius of fresh red meat. The shelf life attainable using conventional vacuum packaging cannot dependably achieve the shelf life needed for export or for distribution to distant markets. Controlled atmosphere packaging using saturating volumes of 100% CO2 with strict temperature control at ÿ1.50.5 C could enable worldwide distribution of fresh meat (Gill & Phillips, 1993; Holley, Delaquis, Gagnon, Doyon & GarieÂpy, 1993). A subset of this technology called master packaging, where retail-ready portions are prepared and wrapped with O2-permeable ®lms and then grouped together and overwrapped in gas impermeable packages, may ®nd application in local and intra-continental fresh meat distribution where labour and space for ®nal product preparation at retail are not available economically (Tewari, Jayas & Holley, 1999). Master packaging successfully combines the bacteriostatic eect created by modi®ed atmosphere packing with CO2 and provides for the oxygenation of the muscle pigment myoglobin (producing the consumer-desired red colour) through the utilization of double-layer packaging. However, a system which can successfully and economically
0963-9969/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved. PII: S0963-9969(00)00079-X
760
M.N.N. Habok et al. / Food Research International 33 (2000) 759±765
maintain the temperature requirements and work in conjunction with master packaging technology has yet to be developed. The objectives of this study were: (i) to modify a container designed by Bailey, Jayas, Holley, Jeremiah and Gill (1997) which uses liquid N2 as the refrigerating medium, to reduce the N2 use of the system, (ii) to test the modi®ed container's ability to chill and maintain meat temperatures over a range of operating conditions, and (iii) to compare the use of a steel braided transfer line for N2 delivery to the container with a vacuum insulated transfer line. 2. Materials and methods 2.1. Design modi®cations and instrumentation 2.1.1. Experimental apparatus A container previously designed and tested by Bailey et al. (1997) was used in this study. The container (Model C-54, Les Contenants Xactics LteÂe, Joliette, PQ) was equipped with a stainless-steel shelving-system, capable of holding 36 master trays at nine levels, with four master trays at each level (Fig. 1). The shelving unit is able to accommodate master trays with dimensions of 50838160 mm, which contain packaged red meat. The liquid N2 refrigerating medium used to chill the container was supplied from a pressurized tank (Dura Series, MVE, New Prague, MN) via a transfer line. Liquid N2 was dispersed in the head space with the use of copper piping and agricultural sprayer nozzles. The design by Bailey et al. relied on the use of six fans loca-
ted in the headspace at the top of the container to circulate the cooling medium throughout the container. Regulation and control of the N2 injections was accomplished through a computer control algorithm (Bailey, 1997) which delivered N2 to the container via a solenoid valve (Model 8263G206 LT, ASCOlectric Ltd., Brantford, ON) located in the transfer line. The container was placed within an environmental chamber (Model C1010, Conviron, Controlled Environments Ltd., Winnipeg, MB) to enable testing of the container over a range of outside temperatures. 2.1.2. Preliminary design modi®cations The heat generated by the fans in the former design contributed to the hold period N2 use of the system. Therefore, this study included an investigation into circulating N2 within the container without the use of fans, using dierent nozzle positions and dierent nozzle types. However, all of the designs tested created temperature gradients within the container and the temperature values were outside the desired range. 2.1.3. Final design of the container A fan was added to enhance uniformity of the temperatures within the container and to improve the accuracy of the temperatures in relation to the target temperature of ÿ1.50.5 C. Bailey et al. (1997) determined that hold period N2 use increased as the number of fans increased, due to heat generated by the fan motors situated inside the container. Therefore, it was decided that the addition of a fan would only prove bene®cial if the additional heat generated by its motor did not enter the container. The resulting fan design
Fig. 1. Schematic of the container used by Bailey (1997) but showing the installation of copper piping and nozzle placement as modi®ed in the present study.
M.N.N. Habok et al. / Food Research International 33 (2000) 759±765
(Fig. 2) was the ®nal fan design for the present system. It consisted of a centrifugal fan encased within a metal enclosure, drawing N2 in from an open bottom. The metal enclosure discharged N2 from the fan into a plenum, consisting of a wooden box, which, in turn, discharged air into two circular (each with a diameter of 100 mm) metal conduits extending vertically downwards from the plenum. The circular conduits were connected to the central conduit within the shelving structure of the container. The central conduit was perforated with a series of 19 mm diameter holes to allow the N2 gas to ¯ow outward from the central conduit over the meat and up along the side conduits (Fig. 2). Nitrogen was delivered to the nozzles within the container with 9.5 mm diameter copper piping and was injected using four V-pattern nozzles (Model 11006VS, Even-Spray and Chemicals Ltd., Winnipeg, MB) which pointed upwards at the level of second layer from the bottom shelf (Fig. 1). 2.1.4. Instrumentation Instrumentation was needed to monitor the temperatures and N2 use of the container. All data were written to the hard drive of a 486 PC by a Quick BASIC program. To measure N2 use, the N2 tank was placed on a ¯oor scale (Model:2136, Mettler-Toledo Inc., Burlington, ON) and its mass was taken every minute by the QuickBASIC control algorithm. An RS 232 connection was used to transfer the mass data to the hard disk. Forty-four thermistors (Model: 44034, Omega Engineering Inc., Stamford, CT) (each with an accuracy of0.1 C) were used to measure the meat temperatures throughout the container and two thermistors were used to measure the temperature of the Conviron chamber. Two thermistors, located inside the container on the second from top and bottom shelves, were used as sensors for the control algorithm. A data acquisition/ control unit (Model 3852A, Hewlett-Packard Co., Loveland, CO) was used to measure the resistance of
761
the thermistors and in turn, relayed the information to the computer through a serial port. Activation of the solenoid valve was controlled by the control algorithm. 2.1.5. Modi®cations to the temperature control algorithm The control algorithm used by Bailey (1997) relied on the concept that the temperature of the product would converge to the cyclic mean temperature of the surrounding air inside the container. However, the meat has to be chilled to the target temperature within approximately the ®rst 2 h of storage to ensure the microbiological safety of the meat. Preliminary testing of the container, following the design changes in this study using pork roasts and simulated master trays, showed that it took approximately 20 h to reach the target temperature. This was because the algorithm would prematurely shut o the solenoid before the meat attained the target temperature. Therefore, the temperature control algorithm was modi®ed to use actual meat temperatures to control the activation of the solenoid in the chilling portion during storage to bring the meat to the desired temperature in the shortest possible time. In addition, the target temperature for the meat was raised to ÿ1.00.5 C to ensure that none of the meat would freeze. During our preliminary testing, some surface ``crusting'' occasionally occurred at ÿ1.50.5 C. Therefore, it was decided that an additional margin of safety should be built into the target temperature of the control algorithm. The formerly used control algorithm was designed so that N2 would only be injected throughout the last 40 s of the 60 s time cycle to allow for an equilisation of temperatures within the container and to allow the instrumentation system to take temperature measurements in the ®rst 20 s of the cycle. However, the data acquisition unit used in this study (Model 3852A, Hewlett-Packard, Avondale, PA) does not require the additional time to take readings, therefore temperature
Fig. 2. Schematic of the ®nal fan design used in this study and the ¯ow of N2 shown by arrows inside the container.
762
M.N.N. Habok et al. / Food Research International 33 (2000) 759±765
readings can be taken simultaneously with N2 injections. Furthermore, if the activation of the solenoid relies solely upon the meat temperature, then an equilisation period is not needed. To ensure that none of the meat would freeze during the chilling period, the minimum meat temperature was not allowed to drop below ÿ1.75 C. In addition, the constraint preventing the air temperature within the container to fall below ÿ20.0 C ensured that the meat would not freeze due to heat transfer from the air to the meat. This value for the minimum air temperature was decided by trial and error. 2.1.6. Simulated master trays Testing of the ®nal container design was conducted with meat samples (three 2 kg pork butt roasts) contained within each simulated master tray. The design of simulated master trays was based on the geometry of an actual master tray so that the ¯ow characteristics of N2 within the container and the heat transfer within the trays could be modelled in a realistic manner. The design of the simulated master tray (Fig. 3) consisted of an aluminum roasting pan (47.336.56.7 cm), with an overlying wire frame, bent to create a head space of air above the 6 kg meat per tray. The tray and frame were enclosed in an unsealed plastic bag (5687 cm) to allow positioning of a thermistor for meat temperature measurement. The head space between the meat and the plastic bag simulated the dead space created within a master package by the back-¯ushed atmosphere. This dead air space aected the heat transport characteristics of the tray in a manner similar to an actual master tray. The thermistor used to measure the meat temperature was passed through the plastic wrap on the meat and was attached to the upper meat surface without penetrating it and was held in place with duct tape. The surface temperature of the meat was chosen as it is the most sensitive indicator of muscle tissue freezing. 2.2. Testing procedures 2.2.1. Temperature and nitrogen use tests The Conviron chamber was used to chill the container and the meat samples to 71.5 C because this was
Fig. 3. Schematic of a simulated master tray.
considered an average temperature of meat entering the chilling process at a processing facility (Gill & Phillips, 1993). This diered from the testing procedure used by Bailey et al. (1997) in which saline water bags, simulating meat samples, were at an initial temperature of 101.5 C. After the initial temperature had been maintained brie¯y (45 min), the container door was closed and the Conviron chamber was then set to the desired outside temperature for the test to be conducted. The liquid N2 release valve of the N2 tank was then opened and the temperature control algorithm was started. The preliminary tests of the container were conducted for 8 h (Bailey et al.), however, tests conducted with the ®nal design of the container lasted for approximately 2 days to ensure that the system had reached a stable temperature. Three replicates were completed for the outside temperatures of 30, 15, 0, and ÿ15 C. 2.2.2. Comparison of vacuum insulated and steel braided transfer lines Two replicates at a chamber temperature of 15 C were conducted to compare the eect on N2 use of using a vacuum insulated transfer line with an uninsulated steel braided transfer line. The latter was used because of the low cost of purchase, its durability, ¯exibility and ease of use during tank installation. Each of these replicates were evaluated for about 24 h. 2.2.3. Failure testing Failure testing was conducted to simulate a N2 failure and to determine the resulting eect on the container temperature. The container and meat samples were equilibrated to ÿ1.01.5 C using the Conviron chamber. After this temperature was reached, the container door was closed and the Conviron chamber was then set to 30 C. Temperatures throughout the container were then recorded for 8 h and three replicates were conducted. 2.2.4. Method of data analysis Temperature history and N2 use versus time were studied using observations at 60 s intervals. The means of maximum, average, and minimum temperatures of meat samples were determined from 5.5 h after initial chilling from 7 C until the trial was completed (Table 1), because the meat samples had reached temperature equilibrium at that point. Nitrogen use in terms of its pre-cool (®xed) and hold (variable) components were determined. The former refers to the quantity of N2 required to chill the contents of the container from 7 C to the desired temperature, while the latter represents the amount of N2 needed to maintain the desired temperature. Pre-cool period N2 use was determined as the amount of N2 consumed up to the beginning of the horizontal section of the curve for N2 use versus time (Fig. 4). Hold period N2 use was
M.N.N. Habok et al. / Food Research International 33 (2000) 759±765
calculated by performing a linear regression on the N2 use data over time from 5.5 h until test completion. The horizontal portion of the curve was not used in calculating N2 consumption. After 5.5 h, the N2 use by the container had reached a constant value (Fig. 4). The slope of the line between N2 use versus time represented the hold period N2 use. Table 2 shows the experimental values for pre-cool and hold period N2 use. 3. Results and discussion 3.1. Temperature and nitrogen use data 3.1.1. Temperature There were no temperature gradients detectable within the container during testing. The temperature range measured in the container was similar for each outside temperature tested. However, it appeared that there was a slight rise in the temperature range as the outside temperature decreased, i.e., the maximum values shifted approximately 0.1 to 0.2 C higher at 0 and ÿ15 C than at the higher outside temperatures. The present design has the additional advantage of being able to operate at sub-zero outside temperatures without freezing of the meat samples because of heat generated by the fan. As the target temperature range for this study (ÿ1.00.5 C) and the previous study by Bailey et al. (1997), (ÿ1.50.5 C), are not the same, a direct comparison of temperature uniformity between the two cannot be made. However, the temperature Table 1 The maximum, average, and minimum temperatures of meat samples within the container from 5.5 to 48 ha Outside temperature Trial number ( C)
Temperature ( C)
Table 2 Experimental values of pre-cool and hold period nitrogen use
30
ÿ1.2 ÿ1.1 ÿ1.1
1 2 3
33.0 30.6 37.4
4.5 5.0 5.0
15
ÿ0.9 ÿ0.8 ÿ0.8
ÿ1.0 ÿ1.0 ÿ1.1
1 2 3
25.2 31.6 38.2
2.8 2.7 2.8
0
ÿ0.5 ÿ0.3 ÿ0.4
ÿ0.9 ÿ0.9 ÿ0.9
ÿ1.3 ÿ1.2 ÿ1.2
1 2 3
38.6 31.2 34.2
1.2 1.2 1.2
ÿ15
ÿ0.3 ÿ0.2
ÿ0.8 ÿ0.8
ÿ1.0 ÿ1.1
1 2 3
33.2 29.6 29.0
0.0 0.1 0.1
Steel braided transfer linea 15 1 2
41.0 38.4
5.5 5.5
1 2 3
ÿ0.6 ÿ0.7 ÿ0.5
ÿ0.9 ÿ0.9 ÿ0.9
ÿ1.4 ÿ1.1 ÿ1.1
15
1 2 3
ÿ0.7 ÿ0.5 ÿ0.4
ÿ1.0 ÿ0.9 ÿ0.9
0
1 2 3
ÿ0.4 ÿ0.4 ÿ0.3
1 2 3
a
Fig. 4. Typical N2 use data from a replicate trial at an outside temperature of 15 C. Pre-cool N2 consumption was calculated using data on the ordinate from 0 to the beginning of the horizontal portion of the curve indicated by an arrow.
Hold period nitrogen use (kg/h)
Minimum
Steel braided transfer lineb 15 1 2
3.1.2. Nitrogen use Table 2 shows the pre-cool period N2 use values for each test and these average values for trials at 30, 15, 0, and ÿ15 C were 33.7, 31.7, 34.7, and 30.6 kg, respectively. Pre-cool period N2 use was similar for all temperatures tested (ScheeÂ's mean comparison test; =0.05). This was expected because the amount of meat was nearly the same for all trials and heat transfer between the container and the outside air would be negligible during the 2 h period of initial chilling. Bailey et al. (1997) observed an average pre-cool period N2 use of 18.1 kg, while the data from this study
Pre-cool nitrogen use (kg)
Average
ÿ15a
range was smaller with the present design than with the previous design (Table 3). Also, the variability in temperature range with a change of outside temperature was more consistent with the present design (Table 3).
Trial number
Maximum
30
763
Trials conducted for 24 h with the same meat used for the three replicates. b Steel braided transfer line as opposed to the vacuum insulated transfer line used in all other tests.
Outside temperature ( C)
a Steel braided transfer line used as opposed to the vacuum insulated transfer line used in all other tests.
764
M.N.N. Habok et al. / Food Research International 33 (2000) 759±765
Table 3 A comparison of the average container temperature ranges for dierent outside temperatures between this study and the study by Bailey (1997) Outside temperature ( C)
Average temperature range ( C) Present study
Study by Bailey (1997)
30 15 0 ÿ15
0.6 0.6 0.7 0.8
1.2 1.0 0.6 0.4
yielded an average value of 32.7 kg. The increased precool period N2 use for the present design is attributable to the fact that Bailey et al. conducted trials using saline bags equivalent to 3.6 kg of meat per master tray, while in this study simulated master trays contained approximately 6 kg of meat per tray. The increased amount of meat within the container used for the present study increased the pre-cool period N2 requirement for initial chilling. Another factor that contributed to the dierence in pre-cool period N2 use was that the initial meat temperatures used by Bailey et al. were 3 C higher than those used here. Hold period N2 use is shown in Table 2 and as expected, the N2 use during the hold period increased as the outside temperature increased. This was due to the higher rate of heat conduction occurring through the transfer line and container walls at the higher outside temperatures. As outside temperature increased, hold period consumption of N2 increased in a proportional manner. Increasing the outside temperature from 15 to 30 C resulted in the N2 use during holding to almost double. Bailey et al. (1997) reported achieving average hold period N2 use values of 5.5, 4.0, 2.6, and 0.9 kg/h for 30, 15, 0, and ÿ15 C outside temperatures, respectively, using the four fan design. In this study, values of 4.8, 2.8, 1.2, and 0.1 kg/h for 30, 15, 0, and ÿ15 C outside temperatures, respectively, were determined for hold period N2 use. Therefore, it can be concluded that the present design represented an improvement in the eciency of hold period N2 use by comparison with the previous design, especially since the storage capacity was almost doubled in the newer design. 3.2. Comparison of vacuum insulated and steel braided transfer lines A comparison of N2 use and container temperatures resulting from installation of either a steel braided or vacuum insulated transfer line was made at an outside temperature of 15 C (Tables 1 and 2). Average pre-cool N2 use was similar for both types of transfer lines (LSD mean comparison test; =0.05) with values of 31.7 and
Table 4 Average rate of meat temperature increase after a nitrogen failure at an outside temperature of 30 C Trial number
1 2 3
Average rate of temperature increase ( C/h) Present study
Study by Bailey (1997)
0.4 0.5 0.5
1.9 2.0 2.0
39.7 kg observed for tests with vacuum insulated and steel braided transfer lines, respectively (Table 2). In contrast, hold period N2 use was signi®cantly aected by the type of transfer line used. Average values for hold period use were 2.8 and 5.5 kg/h, respectively, for the vacuum insulated and steel braided transfer lines. This result was expected because the vacuum insulated transfer line would reduce heat conduction. It was important to quantify changes in N2 use using dierent lines because the ¯exible (braided) line oers convenience and durability during hook up of the tank. Nonetheless, data showed that a vacuum insulated transfer line should be used. A comparison of the data from studies of the two types of transfer lines showed that although the minimum and average temperatures were similar, the tests with the steel braided transfer line had a maximum temperature which was approximately 0.3 C higher than that obtained for the vacuum insulated transfer line. It is unknown why this phenomenon occurred. 3.3. Nitrogen failure A linear regression was performed on the average temperatures of the meat samples during a test to determine the rate of temperature increase following a simulated N2 failure (Table 4). The present system had an average rate of temperature increase following a N2 failure at approximately 1/4 of the values determined by Bailey et al. (1997). 4. Conclusions The following conclusions can be drawn from this study: . Fans were essential to achieve the temperature uniformity required throughout the container (ÿ1.00.5 C). A single fan with the motor mounted outside the container and nozzles injecting N2 upwards along the side conduits at a level near the second-from-bottom shelf, served as the best design in terms of N2 use and temperature uniformity.
M.N.N. Habok et al. / Food Research International 33 (2000) 759±765
. The container operated inside the temperature range (ÿ1.00.5 C) for the average and minimum container temperatures, however, slightly outside the range as far as the maximum temperature is concerned. Maximum and minimum temperature readings for the container were ÿ0.3 and ÿ1.4 C, respectively. . Use of a vacuum insulated transfer line was bene®cial with regard to the N2 use. Use of a vacuum insulated transfer line reduced hold period N2 use almost in half compared to the braided line. . The present design signi®cantly improved the hold period N2 use over the former design by Bailey et al. (1997) and would serve as a more economically feasible method for transport of retail-ready fresh meat.
Acknowledgements The authors thank Messrs. Dale Bourns, Matt Macdonald, and Jack Putnam for their technical assistance and the Natural Science and Engineering Research Council of Canada for partial funding of this project. We thank Dr. C.O. Gill and Dr. L.E. Jeremiah for their technical advice during execution of this research project.
765
References Anon. (1995). Industry pro®le Ð fresh red meat. Report commissioned to Kline & Co. Inc. New York, by AlliedSignal Inc., Morristown, NJ (http://www.asresin.com/plastic/trailer.map). Bailey, C. G. (1997). Design, fabrication, and testing of a returnable, insulated, nitrogen-refrigerated, controlled-atmosphere, shipping container for distribution of fresh red meat. Unpublished MSc thesis, Department of Biosystems Engineering, University of Manitoba, Winnipeg, MB. Bailey, C. G., Jayas, D. S., Holley, R. A., Jeremiah, L. E., & Gill, C. O. (1997). Design, fabrication, and testing of a returnable, insulated, nitrogen-refrigerated shipping container for the distribution of fresh red meat under controlled CO2 atmosphere. Food Research International, 30(10), 743±753. Gill, C. O., & Phillips, D. M. (1993). Assessment of the preservative capabilities of storage and distribution processes. Technical Bulletin 1993-9E. Canada: Research Branch, Agriculture Canada. Holley, R. A., Delaquis, P., Gagnon, J., Doyon, G., & GarieÂpy, C. (1993). Modi®ed atmosphere packaging of fresh pork (Food Research and Development Centre Agriculture and Agri-Food Canada, St. Hyacinthe, PQ). Alimentech, 6(3), 16±17. Shay, B. J., & Egan, A. F. (1990). Extending retail storage life of beef and lamb by modi®ed atmosphere packaging. Food Australia, 42(8), 399±400, 404. Tewari, G., Jayas, D. S., & Holley, R. A. (1999). Centralized packaging of retail meat cuts: a review. Journal of Food Protection, 62(4), 418±425. Taylor, A. A. (1985). Packaging fresh meat. In R. Lawrie, Developments in meat science (pp. 89±113). New York: Elsevier Applied Science Publishers.