1、4672 A Modified Model to Predict Air Infiltration into Refrigerated Facilities through Doorways Donald J. Cleland, Ph.D. Ping Chen, Ph.D. Member ASHRAE Simon J. Lovatt, Ph.D. Mark R. Bassett ABSTRACT This paper describes a modifcation of the model in the ASHRAE Handbook to predict the rate of air in
2、filtration through doors into refrigerated facilities. The model sums the contributions of air tightness (infiltration when the door is closed), air infiltration due to door openings, and the addi- tional air exchange caused by forklqt trafic through the doors. The mod$ed model was developed based o
3、n measured air injiltration rates through rapid-roll and sliding doors (with and without strip-curtain protection) into seven refrigerated warehouses with sizes ranging from 740 to 12,000 m3 (26,000 to 424,000 f$) operating at temperatures ranging from 3 to -22 97 (38 to -7F) for a range of forklist
4、 movement rates. INTRODUCTION Air infiltration can account for more than half the total heat load for refrigerated warehouses and is the main source of frost on air-cooling coils. Most air infiltration is through door- ways, so optimal design and operation of refrigerated facilities and their associ
5、ated refrigeration systems require accurate methods to predict the air infiltration rate. Implications of excessive air infiltration include increased: refrigeration system capital costs related to the higher heat and moisture (frostcondensation) loads, refrigeration system operating costs (particul
6、arly energy), temperature fluctuations that are detrimental to product quality, deposition of frost (ice)/condensation on floors, doors, walls, ceilings, and product (especially near doors) cre- ating hazards to workers, impairing productivity, caus- ing problems with product acceptability, and requ
7、iring periodic removal. Chapter 12 of the 2002 ASHRAE Handbook-Refrigera- tion (ASHRAE 2002) recommends a predictive model based on the equations developed by Gosney and Olama (1975) for fully established airflow through an open doorway. The predicted airflow is adjusted by a door flow factor to acc
8、ount for the fact that airflow is often not fully developed (Dh, the fraction of time the door is open (DJ, and a door protection effectiveness (m. While supplementary equations and guide- lines are provided to select appropriate values of these factors, there are two major weaknesses. First, there
9、is no allowance for air infiltration when doors are closed (air tightness). Such air leakage is likely to be slow but can be significant overall because it occurs for a large fraction of the time. Second, the effect of traffic through a door is taken into account via the door flow factor or the prot
10、ection effectiveness depending on whether the door is protected. For protected doors, the ratio- nale appears to be based on traffic disrupting any protection, thereby reducing E, while for unprotected doors it is implied that traffic prevents airflow becoming fully developed, thereby lowering Df In
11、 both these cases, the air infiltration rate would be expected to be a function of traffic frequency. Unfortu- nately, the guidelines on the effect of traffic are limited to a single trafic frequency of one entry and exit per minute for an unprotected door and a broad range of E values for protected
12、 doors, with little quantitative link to traffic frequency. In this paper experimental data on air tightness and the effect of forklift traffic on air infiltration are presented, and an alternative model for air infiltration into refrigerated facilities based on these data is described. D.J. Cieiand
13、 is a professor and P. Chen is a research assistant in the Institute of Technology and Engineering, Massey University, Palmerston North, New Zealand. S.J. Lovatt is a research leader at AgResearch, Hamilton, New Zealand. M.R. Bassett is a senior researcher in the Build- ing Research Association of N
14、ew Zealand, Porirua, New Zealand. 58 02004 ASHRAE. Table 1. Descriptions of the Refrigerated Warehouses and Doors Tested * RR - rapid-roll door; SD - sliding door; P -personnel door LITERATURE REVIEW The effects of door size on air infiltration, inside and outside air conditions (temperature and rel
15、ative humidity), and door protection (such as air curtains or plastic strip-curtains) for standard slow-acting sliding doors are reasonably well known. Tamm (1 965) derived a theoretical equation to predict the fully developed air interchange rate through a door between a warm area (infiltration air
16、) and a cold area (refrig- erated air). Gosney and Olama (1 975) modified Tamms equa- tion according to equal mass flow rates instead of equal volume flow rates and incorporated a correction factor based on measurements of air infiltration for a model coldstore to get the model recommended by ASHRAE
17、 (2002): air infiltration rate for fully developed flow (L/s), door width (m), door height (m), gravitational acceleration (m/s2), ratio of infiltration air density to refrigerated air density =pi /p, density of infiltrated air (kdm3), density of refrigerated air (kg/m3). Equation 1 predictions are
18、about 60% to 70% of those given by the equation derived by Tamm for typical tempera- ture differences across a door. Longdill et al. (1974) validated Tamms equation experimentally for a 1.2 m (3.9 ft) wide by 1.6 m (5.2 ft) high doorway in a 177 m3 (6250 ft3) coldroom. Pham and Oliver (1 983) and Fr
19、itzsche and Lilienblum (1 968) measured air infiltration for a range of doors in refrigerated warehouses. For unprotected doors they found that Tamms equation was a good predictor for narrow door widths and short door opening times (53 minutes), but otherwise it consis- tently overpredicted the tota
20、l air infiltration rate by about 20% to 50% (i.e., measured rates were similar to those that would be predicted by Equation 1). Hendrix et al. (1989) found that Equation 1 gave good predictions of air infiltration for fully developed flow, and that steady-state flow becomes estab- lished 3 s after t
21、he door opens, whereas Azzouz and Duminil (1 993) found flow development lag times of less than 2 s. The effectiveness of door protection devices such as plas- tic strips and air curtains is defined as the fraction by which air infiltration is reduced compared with an unprotected door. Air curtains
22、have been found to typically be 49% to 83% effective (Longdill et al. 1974; Pham and Oliver 1983; Downing and Meffert 1993). The effectiveness of plastic strip-curtains has been variously reported as 86% to 96% without forklift traffic (Pham and Oliver 1983; Hendrix et al. 1989; Downing and Meffert
23、1993) and 82% to 92% with one forklift entry and exit per minute (Downing and Meffert 1993). Pham and Oliver (1983) also investigated the effect of forklift trafic on air infil- tration at a passage frequency of one entry and exit per minute. They found the infiltration rate decreased by 21+10% for
24、unprotected doors and increased by 32I45% for doors protected by plastic strip-curtains compared with doors being opened without trafic. More recently rapid-roll and fast-folding doors have been used, especially in warehouses with high frequency of forklift movements. Claimed advantages include auto
25、matic opening and closing, fast action, reduced open times, good air sealing, and good thermal resistance. The effectiveness of fast-acting doors tested by Downing and Meffert (1993) was about 93% (similar to strip-curtain effectiveness) without trafic (Le., when closed) and 79% to 85% with trafic a
26、t a rate of one entry and exit per minute. The doors were fully open for 8 to 20 seconds and took 1 to 2 seconds to open or close for each traffic pass. Presumably most of the benefit arose due to the reduced open time rather than lower infiltration rate when open, although the lag effect described
27、above may have contributed. ASH RAE Transactions: Research 59 Site A5 A6 Unfortunately, few other independently measured data and theoretical analyses of air infiltration through warehouse doors, especially the effect of trafic movement, are available. Also, there is little information on the signif
28、icance of air infil- tration into refrigerated buildings when doors are closed (airtightness). Door Action Times (s) Total Equivalent Open Time (averageIstd.dev.) Per Movement* (s) Opening (a) Fully Open (in) (bin) Fully Open (out) (bout) Closing (c) . (a + bin+ bout+ c) 1.7kO.l 17.0k1.0 15.7fl.2 3.
29、1fO. 1 37.5+ 2.4 2.910.1 15.4k8.6 8.8k2.8 4.9k0.1 32.0Il1.6 EXPERIMENTAL PROCEDURES Measurements of air infiltration were carried out in the seven typical single-story refrigerated warehouses summa- rized in Table 1. For each warehouse, only the main forklift door was used for loading/unloading prod
30、uct during the trials. All other doors (e.g., personnel doors) were kept tightly closed. In all cases, the product bins or pallets being stored and moved by forklifts through the doors were similar in size (about 1.1 m 3.6 ft long, 1 .O m 3.3 ft wide, and 2.0 m 6.6 ft high). Air infiltration was mea
31、sured using the tracer gas decay technique described by ASTM (1983) using SF, as the tracer gas. Full details of the methodology are given by Chen et al. (1999, 2002). Two warehouses (A5 and A6) had rapid-roll doors (RRD) with static plastic strip-curtain (SC) protection. The polypro- pylene door fa
32、bric opened by rolling vertically upward, while the strip-curtains were attached to the door frame inside the warehouse. These doors were automatically activated by magnetic sensors mounted above and beside the door. The operating characteristics of both rapid-roll doors were measured for over 40 pr
33、oduct movements by forklift under normal operating conditions. Table 2 summarizes the results. The other warehouses had one or more slow-acting slid- ing doors (SD) operated manually via a pull-cord. All sliding doors except one had static plastic strip-curtain protection. The time for each sliding
34、door to open or close was consider- ably longer than for the rapid-roll doors and reasonably consistent for each door (between 4 and 10 s). However, the time that each sliding door remained fully open for each fork- lift movement was highly variable. Sometimes the doors were closed after entry and r
35、eopened when the forklift exited the warehouse; sometimes the doors remained open for the whole product movement because it was not worth closing while the forklift was inside; and in other cases the door was kept open for a series of product movements or permanently. For sliding doors being separat
36、ely opened for each product movement, the fully open time was typically about 60 s. Extensive trials were performed on the rapid-roll doors (warehouses A5 and A6). The trials were carried out at off- peak times to minimize their effect on normal operations of the warehouses and to allow the door-ope
37、ning frequency and duration to be artificially controlled if required. The regimes tested were: (a) door fully closed (i.e., airtightness); (b) door fully open with and without a strip-curtain, without forklift traffic; (c) door fully open for different durations with similar frac- tional open time
38、(e.g., open 3 s every 15 s, open 10 s every 50 s, open 20 s every 100 s, or open 30 s every 150 s) without a strip-curtain; (d) door fully open for different fractional open times, in a repeating cycle (e.g., open for 10, 20, 30,40, or 50 s in a 100 s cycle) with and without a strip-curtain, without
39、 forklift traffic; (e) forklift traffic passing through the door at different fre- quencies (1 to 75 movements per hour) with and without a strip-curtain, with the door on automatic sensor or pull- cord control or permanently open; (f) door open at different frequencies to mimic forklift traffic but
40、 without a forklift passing through, with and without a strip-curtain. For the other doors the full set of regimes could not be performed due to warehouse operating restrictions. The door in warehouse Al was only monitored without a strip-curtain and without traffic. The doors in warehouses A2a, A2b
41、, A4, and A7 were not monitored without their strip-curtain protec- tion. For some of the sliding doors, it was possible to artifi- cially control door open times as part of regimes (e) and (f). In all warehouses, the fans suctions and ducting were located away from the doors so they were not expect
42、ed to affect air infiltration rates. Some regimes were replicated at least three times so that variability could be assessed. RESULTS AND DISCUSSION Measured air infiltration rates for the various warehouses, doors, and operating conditions and regimes are summarized in Table 3. Airtightness Airtigh
43、tness depends mainly on the quality of door seals and any flow through pressure equalization ports. If pressure equalization ports are appropriately located (away from fan 60 ASHRAE Transactions: Research Table 3. Measured Airtightness, Door Flow Factor, Strip-Curtain Effectiveness, and Air Infiltra
44、tion Due to Forklift Traffic A2a A2b A3 A4 A5 A5 A6 A6 A7 A7 A7 -0.4 (3 1.3) -14.6 (5.7) -1.0 (30.2) -16.2 (2.8) (37.9) (-1.1) 3.3 -18.4 -21.8 (-7.2) -13.9 (7.0) 30.9 (55.6) 15.4 (27.7) (66.2) 36.8 42.3 (76.1) 32.3 (58.1) 12 0.4 0.88 0.87 43300 5200 Small gaps in SC (25) (0.3) (1530) (180) Enclosed
45、door, DPO 20 0.4 15600 5200 DPO, with SC (42) (0.3) (550) (180) (127) (1.8) (1 500) (170) (630) 60 2.8 1.27 0.92 42500 4800 With SC -54200 17700 Without SC (- 1 9 1 O) 90 4.0 1.31 0.80 67600 5 1 O0 Poor door seal (1 5 md0.6 (191) (2.6) (2390) (180) in. gap) One strip missing in SC 66 2.9 Gap at bott
46、om covered (140) (1.9) With SC 34 1.5 36800 6100 Poor door seal (20 md0.8 (72) (1 .O) (1300) (220) in. gap) With SC 41400 6800 Poor door seal (20 md0.8 (1460) (240) in. gap) DPO, with SC 27 1.2 Gap at bottom covered (57) (0.8) With SC * SC - strip-curtain; DPO -door permanently open suction and disc
47、harge areas, protected from prevailing winds, and at the appropriate height), and the warehouse temperature does not vary substantially from the setpoint temperature, then the potential for airflow through the port due to pressure differ- ences is low. In such cases, it is expected that the main fac
48、tors influencing airtightness will be the seals around door. The driving force for airflow through seals is hydrostatic air pres- sure differences, so the difference between refrigerated and infiltrating air temperature and, hence, air density is also likely to be important. The results given in Tab
49、le 3, indicate that the air leakage per length of door seals (QI Hendrix et al. 1989; Downing and Meffert 1993). However, the effectiveness of a strip-curtain with one strip missing (A6) was only 80%, which means that the infil- tration rate is nearly three times higher than if the curtain was in good condition. Similarly, the E of a strip-curtain with small gaps between strips was only 87%, which corresponds to about 1.5 times the infiltration rate for a curtain in good condi- tion. These results highlight the benefits of careful strip- curtain installation and regular mainten
