Technical Assessment of Dry Ice Limits on Aircraft (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

26 package was filled with dry ice pellets, the top was closed and taped, and the package was placed on the scale. The weight was recorded manually at intervals along with the time and, in some tests, one or more inside temperatures. The test packages were held in an air-conditioned room with a nominal temperature of 22°C to 24°C during the tests. Some additional tests were made to determine the density of the EPS foam and the thermal resistance of the cardboard. Results Some representative results for the laboratory package tests are shown in Figure 6. The area-normalized sublimation rates were in the range of 130 to 180 g/m2 ? hr. Note that the mass of dry ice declines smoothly with time and the area- normalized mass loss is approximately constant. Several of the test conditions were repeated using about half the normal loading of dry ice in the package; the carbon dioxide emission rate measured in kg/m2 ? hr did not change appreciably. Laboratory Tests of Multiple Package Configuration Objective The objective of the multiple-package tests was to gain insight into the effect of neighboring packages on the dry ice sublimation rate. Test Configurations Two basic test configurations were studied. In both con- figurations, the test package was resting on a plywood base (to simulate the top of a wooden pallet) and was surrounded by packages of the same dimensions on all four sides and on the top. The adjacent packages were touching the test pack- age but were not held in place by bands, shrink wrap, or In addition to calculating sublimation rates from heat transfer rates, both laboratory and field tests were performed to measure sublimation rates experimentally. These results are reported in the following. The objective of these experimen- tal package tests was to determine the observed sublimation rate under a variety of experimental circumstances. Laboratory Tests of Single Packages Objective The objective of the laboratory tests was to establish the dry ice loss rate of packages under controlled conditions. Test Packages The test packages were EPS-insulated cardboard cartons of three different sizes. Basic information about these packages is provided in Table 6. Mass and Temperature Measurement The scale used for these tests had a full-scale range of 15 kg, a readability of ±2 g, and repeatability and linearity of ±5 g. The scale was zero-checked, then calibrated at mid-scale and near full-scale with calibrated weights; the calibration of these weights is traceable to the National Institute of Standards and Technology (NIST). The scale pan was approximately 210 mm square. The cross-sectional dimensions of all the packages were greater than 210 mm. Thus, part of the bottom surface of the package was in contact with the scale pan, and the balance of the area was surrounded by air. The internal temperature measurements were made using type E thermocouples con- nected to a Fluke 52II thermocouple temperature meter. Procedure The procedure used for the package weight loss tests was simple: the tare weight of the package was determined, the C h a p t e r 8 Experimental Measurements of Dry Ice Sublimation Rates

27 other constraints. It should be noted that although all the packages were new and unused, the corrugated cardboard used to construct the packages was not perfectly flat, and not all the packages were perfectly square. We believe this situation would be typical of packages actually shipped in the field. For this test configuration, the surrounding packages con- tained EPS foam lining but were otherwise empty. These packages were intended to simulate general merchandise, which is often surrounded by foam- or paper-based cushion- ing to prevent breakage. Results As expected, the dry ice sublimation rate was lower for a package surrounded by other packages and therefore not Table 6. Test package information. Package Designation Width x Depth x Height, mm Surface Area, m2 Insulation Thickness, mm Tare Weight, kg A 300 x 240 x 390 0.56 38 0.70–0.75 B 290 x 240 x 200 0.35 38 0.40–0.45 C 370 x370 x 390 0.85 38 1.1 Figure 6. Representative package test results. % 0 % 0 1 % 0 2 % 0 3 % 0 4 % 0 5 % 0 6 % 0 7 % 0 8 % 0 9 % 0 0 1 8 4 2 4 6 3 0 3 4 2 8 1 2 1 6 0 d e s p a l E , e m i T r h y r D e c i , t f e l % n o i t a m i l b u S , e t a R r h / % X 0 1 PACKAGE TEST 11 BOX A003

28 exposed to the free air convection and radiant heat exchange that would obtain for a single package for which these pro- cesses are effective means of heat transfer. For this configura- tion, the measured area-normalized sublimation rates were in the range of 130 to 160 g/m2 ? hr. Package Shipment Tests Procedure In these experiments, the same type of EPS-insulated ship- ping cartons used in the laboratory tests were filled with dry ice pellets and shipped between two Battelle offices: the Battelle headquarters in Columbus, Ohio, and a field office near Phoenix, Arizona. The packages were shipped via normal FedEx next-day service. Otherwise the procedure was similar to that for the labora- tory tests. Two scales of the same make and model were used to weigh the packages. The scale used in Phoenix was compared to the one used in Columbus using the same calibration weights, and the two scales were found to read the same within the specified ±2 g repeatability. Results Some representative results for the dry ice package ship- ment tests are shown in Figure 7. Note that the first weight loss measurement data point could not occur until the package was received in Phoenix, about 24 hours after the package was packed with dry ice for shipment. The break in the curves at about 100 hours reflects the depletion of the dry ice pellets. In practice, shipment times would likely be less than 100 hours, and for a lesser amount of time the area- normalized subli- mation rates are comparable to those obtained during the laboratory tests. 0 0 0 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 6 0 0 7 0 0 8 0 0 9 0 0 0 , 1 % 0 % 0 1 % 0 2 % 0 3 % 0 4 % 0 5 % 0 6 % 0 7 % 0 8 % 0 9 % 0 0 1 0 4 1 0 2 1 0 0 1 0 8 0 6 0 4 0 2 0 N or m al iz ed L os s Ra te d e s p a l E , e m i T r h y r D e c i , t f e l % n o i t a m i l b u S , e t a R r h / % X 0 1 d e z i l a m r o N a e r A , s s o L 2 m / m g r h PACKAGE TEST 17 BOX A004 Figure 7. Representative results from package ship tests.

29 Discussion of Experimental Test Results General Observations In Figure 6 and Figure 7, the weight loss with time is a smooth downward curve. Both these figures show the nor- malized area loss as a relatively constant number between 140 and 170 g/m2 ? hr. In Figure 7 we also show the apparent mass-based sublimation rate, expressed as %/hr. Note that the mass-based sublimation rate is not very constant. Uncertainties in Experimental Measurements Regarding the uncertainties in the experimental data, the observed steady-state loss rate from the experiment is believed to be quite accurate. The scale was calibrated using a weight traceable back to NIST. The display was accurate to ±2 g, and because the mass loss was measured over a period of several hours, the mass loss was great enough that this inaccuracy did not significantly affect the measured loss rate. The sublimation of dry ice is believed to be the only cause of a weight change on the scale. No condensation was observed on the outside of the boxes, and because the dry ice is con- stantly subliming, there is a constant carbon dioxide gas flow outward through any gaps between the package lid and the top that would limit any diffusion of moist air into the pack- age and minimize any weight gain from the buildup of water ice crystals on the dry ice pellets. Comparison of Single- and Multiple-Package Test Results Although the area-normalized dry ice loss rates for the multiple-package ensembles were lower than for single pack- ages, the reduction was not extreme—on the order of 20%. Comparison with Heat Transfer Analysis When the dry ice sublimation rates observed in the pack- age tests were compared with the results of the heat transfer analysis presented in Chapter 7, it was found that there was good agreement. The heat transfer rates, in watts per square meter, which were calculated for the packages (based on 38 mm of EPS foam insulation and free convection on the outside of the package), were within 20% of the actual heat transfer rates derived from the laboratory measurements of the dry ice loss rate. Proposed Use of an Area-Normalized Dry Ice Loss Rate The dimensional analysis, the heat transfer model, and the dry ice use rate tests all support the use of an area-normalized dry ice loss rate. We therefore propose that the principal metric for the assessment of the carbon dioxide generating potential of both packages and ULDs carried on aircraft be a normalized area loss rate (mass of dry ice lost per unit of surface area per unit of time).* If this were part of the protocol to establish dry ice limits, the carrier would have to be provided with the dimen- sions of the dry ice packages being loaded on the plane. We think that this is reasonable because in order to assess ship- ping charges, carriers such as FedEx, UPS, and USPS already check package dimensions in order to compute their dimen- sion weight, and because ULDs come in standard sizes whose dimensions are well-known and whose area could even be stenciled on the unit. The graph in Figure 8 presents the results of the pack- age tests along with vendor information on insulated ULDs. Please note the following comments related to the graph: The data for the cartons come from tests conducted by Bat- telle. Because the results for the shipped packages were about the same as those for the package tests in the lab, they have been grouped together. The data for the ULDs come from vendors and have not been independently verified. The package and ULD sizes cover a surface area range from 0.3 m2 to 32 m2, a factor of over 100 from the smallest to the larg- est, and cover the entire range of sizes likely to be seen on aircraft. The temperature inside of the cartons* was near dry ice tem- perature, leading to a DT of about 100 K. The set point for the insulated ULDs (which have active temperature control) was taken at -20°C for a DT of 40 K.† So, the cartons and the ULDs do not have the same DT. However, it appears that, because of the need for an aluminum support frame in the ULDs, even though *In SI units, this would be grams or kilograms of dry ice per square meter per hour. *Based on measurements performed in this task. †Many insulated ULDs are set to maintain interior temperatures of 3°C to 8°C, not -20°C. However, these ULDs are advertised to work at -20°C, and the manuals give the dry ice loadings required to maintain -20°C. The air carrier does not inspect the ULD control panel to check the temperature setting prior to loading, and the crews are not advised of the temperature setting. So, if the basis for higher dry ice limits for ULDs is a higher 3°C to 8°C set point, then (a) the shipper should cer- tify the temperature setting, (b) the airline should have a way to verify the setting, and (c) regulations and policies may need to be rewritten to include a temperature setting requirement. And then there would need to be another procedure for those ULDs whose contents do need to be deep frozen at the lower -20°C temperature. Some pharmaceuticals do need to be deep frozen.

30 the DT for a typical ULD package is about half the DT for a typi- cal dry ice carton, the overall thermal conductivity of the ULD is about twice the thermal conductivity of the EPS in a typical dry ice package, and the two factors cancel out. If a ULD were main- tained at the -78°C sublimation temperature of dry ice, then an adjustment would need to be included in the correlation. This is very unlikely because even deep-frozen cargo only requires a temperature set point of -18°C or -20°C.* The correlation shows that the normalized loss rates, except for the FAA data point, are in the range of about 120 to 210 g/m2 ? hr, with an average of about 170 g/m2 ? hr. For a package with a DT of 100 K, an insulation thickness of 38 mm, and a typical dry ice loss rate of 170 g/m2 ? hr, the sublimation number may be calculated as follows: Sublimation Number = ∆ = ∆ = kL T t r T R rs s a 2 3 4 λ λ . While some additional analyses might be required to demon- strate the validity of these findings, it could be reasonable to require the design of dry ice packages to have a sublimation number of 3.4. Figure 8. Normalized dry ice loss rates for various cargos. 0 0 5 0 0 1 0 5 1 0 0 2 0 5 2 0 0 3 0 5 3 0 0 4 0 0 1 0 1 1 1 . 0 Dr y Ic e Lo ss R at e, g m /m 2 h r e c a f r u S a e r A , m 2 Normalized Dry Ice Loss Rate vs. Surface Area t s e T A A F s e x o B s N K R s P A R s n o t r a C P Y J N K R d n a s t n i o p P A R t n e s e r p e r o w t . s r o d n e v t n e r e f f i d The result of the 2006 FAA sublimation rate test† is also shown. It is believed that this data point is higher than the others because: (1) According to the report, the boxes were initially weighed immediately after filling with dry ice, so the loss rate data would include the time when they were in a cool-down mode. In actual practice, packages would need to be transported from the shipper to an airport (which takes time), and in any case, airlines will not accept freight less than a minimum of 1 hour (domestic) or 2 hours (international) prior to flight time, so in actual practice, cool down would not be an issue. (2) Other factors could also contribute, including (a) the FAA test used relatively small dry ice pellets (10 mm in diameter by 20 mm in length), (b) the lower pressure in the altitude chamber leads to a dry ice equilibrium sublimation temper- ature about 4°C lower than at atmospheric pressure, leading to a DT and associated expected heat transfer rate about 4% larger, and (c) the chamber temperature of 29°C is higher than the ambient temperature used for the other tests (and is higher than typically seen inside an aircraft cargo compart- ment during flight). †As reported in DOT/FAA/AM-06/19. *A company that ships pharmaceuticals reported that they aim to keep their deep-frozen products at -14°C during shipping.