ENGINEERING RESEARCH INSTITUTE THE UNIVERSITY OF MICHIGAN ANN ARBOR ANALYSIS OF THE PERFORMANCE OF FIVE FORCED-AIR COOLING UNITS AT THE DENTON OIL FIELD PLANT OF THE ATLANTIC REFINING COMPANY Report No. 38 Edwin Ho Young Assistant Professor of Chemical Engineering Dennis J, Ward Luis 09 Gonzalez Research Assistants Project 1592 WOLVERINE TUBE DIVISION CALUMET AND HECLA, INCo DETROIT, MICHIGAN December 1955

ACKNOWLEDGEMENT The University of Michigan research group wishes to acknowledge the excellent cooperation and assistance of Mr. Robert H. Fritz in conducting the field tests at the Denton oil field gasoline plant at Lovington, New Mexico.

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE OF CONTENTS LIST OF TABLES LIST OF ILLUSTRATIONS ABSTRACT I. DESCRIPTION OF UNITS AND TEST PROCEDURE A. INTRODUCTION B. UNITS TESTED ON THURSDAY, APRIL 28, 1955 C. UNITS TESTED ON MONDAY, MAY 2, 1955 D. DESCRIPTION OF TEST-METHODS AND PROCEDURE E. DESCRIPTION OF UNITS NOT TESTED II. FIELD TEST DATA ON JACKET WATER COOLER A. DATA B. PRELIMINARY ANALYSIS OF THE DATA Co DISCUSSION OF INCONSISTENCIES D. RESOLUTION OF INCONSISTENCIES E. REVISED ANALYSIS OF THE FIELD TEST DATA III, HEAT TRANSFER CORRELATIONS FOR TUBE BANKS A, CORRELATIONS FOR AIR-SIDE COEFFICIENTS FOR THREE-INLINE, FOURTH-ROW-STAGGERED UNITS Bo TRIANGULAR-PITCH CORRELATIONS IV. PERFORMANCE OF THE THREE-IN-LINE, FOURTH-ROW-STAGGERED UNIT COMPARED WITH THE PERFORMANCE OF THE TRIANGULAR-PITCH UNIT V. HORSEPOWER AND AT CONSIDERATIONS VI. PRESSURE-DROP COMPARISONS VII. FAN PERFORMANCE VIII. LABORATORY TESTS OF ATLANTIC TUBE IX. JACKET WATER TEST-ANALYSIS CONCLUSIONS X. HAPPY CO. TEST DATA Page v vi viii 1 1 1 2 2 4 5 5 10 15 16 17 24 24 39 49 51 55 57 60 64 65 iii

L - ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN - TABLE OF CONTENTS (Concl.) Page XI. HAPPY CO. SPECIFICATION SHEETS 68 XII. ANALYSIS OF UNITS NOT TESTED 74 A. LEAN OIL COOLER 74 B. DEBUTANIZER OVERHEAD CONDENSER76 C. EVAPORATOR CONDENSER 78 D. STILL OVERHEAD PARTIAL CONDENSER79 E. CONCLUSIONS 82 APPENDIX A 83 APPENDIX B 9 APPENDIX C 99 APPENDIX D 104 APPENDIX E 112 APPENDIX F 114 APPENDIX G 121 APPENDIX H 123 APPENDIX I 126 APPENDIX J 135 APPENDIX K 136 APPENDIX L 38 138 - i iv

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN LIST OF TABLES Table Page I SUMMARY OF EXPERIMENTAL DATA OF THURSDAY, APRIL 28, 1955 6 II SUMMARY OF EXPERIMENTAL DATA OF EAST BAY, DATA OF MONDAY, May 2, 1955 8 III SUMMARY OF EXPERIMENTAL DATA OF WEST BAY, DATA OF MONDAY, MAY 2, 1955 9 IV TABULATED RESULTS 21 V COMPARISON OF PREDICTED COEFFICIENTS WITH TEST RESULTS FOR THREE-IN-LINE, FOURTH-ROW-STAGGERED UNIT 39 VI SUMMARY OF TRIANGULAR-PITCH ANALYSIS FROM AIR-SIDE DATA 47 VII COMPARISON OF TRIANGULAR PITCH AND THREE IN LINE, FOURTH ROW STAGGERED 49 VIII HAPPY CO. TEST DATA OF THURSDAY, APRIL 28, 1955 65 IX HAPPY CO. TEST DATA OF MONDAY, MAY 2, 1955 67 X SUMMARY OF PRESSURE-DROP ANALYSIS 55 XI COMPARISON OF PRESSURE DROPS WITH PREDICTED VALUES FOR 600-FT/MIN FACE VELOCITY 56 XII PREDICTED PERFORMANCE OF LEAN OIL COOLER 74 XIII LEAN-OIL-COOLER ANALYSIS BASED ON 651C AND Vrace OF 760 STD FT/MIN 75 XIV LEAN-OIL-COOLER ANALYSIS BASED ON HAPPY CO. RATING CURVE AND Vface OF 760 STD FT/MIN 75 v

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN LIST OF ILLUSTRATIONS Figure Page 1 Jacket water cooler layout. 3 2 Calibration of a four-inch vane-type anemometer in a six-inch-diameter duct. 18 3 Wolverine tube data sheet 651A (outside coefficient). 25 4 Wolverine tube data sheet 651B (outside coefficient). 26 5 Wolverine tube data sheet 651C (outside coefficient). 27 6 Wolverine tube data sheets 651A, 651B, and 651C. (outside coefficient). 28 7 Wolverine tube data sheet 651A (air-side pressure drop). 30 8 Wolverine tube data sheet 651B (air-side pressure drop). 31 9 Wolverine tube data sheet 651C (air-side pressure drop). 32 10 Wolverine tube data sheets 651A, 651B, and 651C (air-side pressure drop). 33 11 Heat transfer and pressure drop.(ALCO Co. data). 36 12 The Happy Co. heat-transfer and pressure drop rating curves, 38 13 Heat transfer correlation for air. 41 14 Generalized heat transfer correlation. 42 15 Heat transfer correlation for air as a function of S, N, and Dr. 43 16 Pressure-drop correlation for air. 44 17 Empirical pressure-drop correlation for air. 45 18 Generalized pressure-drop correlation. 46 19 Heat-transfer correlation of T. E. Schmidt applied to air. 48 20 Outside coefficients vs theoretical horsepower. 52 vi

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN LIST OF ILLUSTRATIONS (Concl.) Figure Page 21 Jacket water cooler anticipated fan performance curves. 58 22 Predicted blade efficiency of jacket-water-cooler fans. 59 23 Results of tests conducted on tubes cut from the jacket water cooler. 61 24 Predicted performance of tubes tested in field application. 63 25 Gardner's fin efficiency. 116 L vii

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN ABSTRACT This report contains an analysis of field test data taken on a jacket water cooler at the Denton oil field gasoline plant of the Atlantic Refining Co. Analyses of the proposed designs for a lean oil cooler, a debutanizer condenser, an evaporator condenser, and a still overhead partial condenser for the same plant are also included. All the above units use air under forced convection as the cooling medium, with the air-side coefficients controlling the performance. The rating curve used by the designers greatly overrates the air-side coefficients, resulting in underperformance of the units. viii

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN I. DESCRIPTION OF UNITS AND TEST PROCEDURE A. INTRODUCTION The Denton oil field gasoline plant of the Atlantic Refining Co. had been on stream for about a year. During this period the plant production capacity was found to be considerably below minimum acceptance during the summer months. The cause of this reduced capacity was traced to several aircooled units which were fabricated from Wolverine bimetal high-fin tubes, The most serious unit involved the cooling of recirculated compressor-jacket cooling water. Field tests were made in order to obtain data which could be used to evaluate the performance of the jacket water cooler. The other field units involved were not tested but the specifications for these units were analyzed. This report contains a summary of the analysis of all the units involved. B. UNITS TESTED ON THURSDAY, APRIL 28, 1955 The tests conducted on Thursday, April 28, were made on the jacket water cooler. This unit is an air-blast heat exchanger using forced air on the outside of the tubes to cool water flowing inside the tubes. The fans are located upstream from the tube bank. The jacket water cooler tested consisted of two separate bays, the east bay and the west bay, each capable of operating independently from one another. Each bay contained 358 Wolverine Trufin tubes (Catalog No. 620916049-26), 24 ft long. Each bay had its own 16-ft, 4-blade, Moore pressureblower-type fan for pushing the air vertically upward through the tube bank. The east bay fan-tip clearance had been reduced by representatives of the Happy Co. prior to the tests, whereas the west bay fan-tip clearance had not been adjusted for these tests. The tubes in both of the two bays were arranged with the bottom three rows in line and the fourth row staggered. A section of one tube had been removed from the west bay and shipped to The University of Michigan for testing. The results of thesie tests are summarized in Section VIII of this report. 1

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN C. UNITS TESTED ON MONDAY, MAY 2, 1955 From April, 29 to May 1, 1955, the gasoline-plant operations were shut down. During this time new tube banks fabricated by the Happy Co. were installed in place of the tube banks tested on the previous Thursday. The fan arrangements were not changed. The new east bay of the jacket water cooler contained the same three-in-line, fourth-row-staggered arrangement of the tubes as the tube bank it replaced. The tube arrangement of the new west bay tube bank was changed to a triangular pitch arrangement. D. DESCRIPTION OF TEST METHODS AND PROCEDURE The University of Michigan research group, assisted by Mr. Robert Fritz of Wolverine Tube, measured the following variables: (a) temperature of the inlet water; (b) temperature of the outlet water; (c) water flow rate; (d) air inlet temperature to the fans; (e) outlet air temperature profile; (f) outlet air velocity profile; and (g) fan-motor amperage. An attempt was also made to measure the static pressure profile between the fan and tube bank; however, it was unsuccessful for reasons discussed in Section II-A of this report. The water inlet temperatures and water outlet temperatures were measured with mercury-in-glass thermometers. The inlet air temperature and the outlet air temperature profiles were measured with iron-constantan thermocouples. The thermocouples used in the outlet air temperature profile were centered in six-in.-diameter stovepipe ducts approximately one ft high. This was necessary in order to avoid mixing of the air coming from the tube bank with the ambient air above the bank. The EMF readings were obtained with a potentiometer furnished by the University of Michigan Engineering Research Institute. The air velocity profile of the air leaving the units was determined by using a Taylor four-in., Model 3132, vane-type anemometer placed concentrically in a six-in. duct; the anemometer time period was measured with a stop watch. The electrical current flowing through the fan motor was measured, using a General Electric hook-on volt-ammeter, Type AK-1. A discussion of the reliability of the test measurements is given in Section II-C of this report. The procedure followed in testing a unit was as follows. Ten thermocouple assemblies were arranged on top of the unit in a grid form as indicated in Fig. 1. The grid areas are identified by letter-number combinations such as 2D, 3A, etc. One thermocouple assembly was placed directly under the fan. Water-thermometer, thermocouple, and the cold-junction temperature readings were obtained in succession until reasonably steady-state conditions were reached. The local barometric pressure was obtained by telephoning the local airport. The velocity profile was made over the same areas 2

ENGINEERING RESEAR < I 0,(1) 3: a I 1 KI I- 1 CH INSTITUTE * UNIVERSITY OF MICHIGAN 0 10 rm 0.,0 0 i* -Po L I I Ia - Z I IT 0 4 — V) 0 L ~ I I I l () 3

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN occupied by the thermocouples during the period when the thermocouple readings were being made. E. DESCRIPTION OF UNITS NOT TESTED In addition to the jacket water cooler there are four other aircooled units on stream at the gasoline plant. These units are identified as (a) Lean oil cooler, (b) Still overhead partial condenser, (c) Evaporation condenser, and (d) Debutanizer overhead condenser. The specifications for these units are given in Section XI of this report. Evaluations of these units are presented in Section XII of this report. 4

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN II. FIELD TEST DATA ON JACKET WATER COOLER A. DATA The field test data obtained on the jacket water cooler are summarized in Tables I, II, and III. Table I presents the test data obtained on Thursday, April 28, 1955. These data were obtained on the unit during a period of fluctuating ambient-air temperatures and under fan-blade pitch control conditions such that fluctuations in the fan pitch resulted in varying throughput of air. Since these test data were not obtained under steady-state conditions, some questions arise as to their validity for determining the performance of the unit prior to the turnaround. Table II summarizes the data obtained on the east bay of the jacket water cooler on Monday, May 2, 1955, for the period indicated. This bay contained 179 tubes on a three-rows-in-line, fourth-row-staggered pitch arrangement. This unit replaced one of the bays tested on the previous Thursday 5

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE I SUMMARY OF EXPERIMENTAL DATA OF THURSDAY, APRIL 28, 1955 Arrangement of Tubes: three rows in line, fourth row staggered Data on West Bay 446.5 58.6 1A L 1B R 1B L 1C R 1C L lD R 1D L 1E R 1E L 2A R 2A L 2B R 2B L 2C R 2C L 2D R 2D L 2E L 2E R 1.48 1.60 1.48 1.50 1.47 1.50 1.57w1.37 1.60 1.47 1.42 1.52 1.60 _.m 1.58 1.60 1.45 1.47 1.47 1.48 1.38 1.60 1.45 1.47 1.57 1.6o _ _ 1.43 1.60 1.465 1.485 1.47 1.49 1.575 1.60 1.46 1.445 1.545 1.60 351.5 494 575 590 419 559.5 623 496 549 366 598 466 258 546 450 551 415 464 377 58.9 58.5 60.1 58.2 62.8 58 59.5 60.0 59.9 59.8 60.0 59.1 60.0 60.0 60.2 59.1 60.0 59.5 60.2 456 358 508 575 608 400 558 628 496 550 367 398 475 258 546 429 539 413 467 376 Cold-Junction Temperature Cold-Junction Temperature (Cell No. 1) (Cell No. 2) = 27.8o0c = 26.2~C West Bay Avg = = 82.0~F = 79.2~F 470.1 Water Inlet Temperature = 163~ Water Outlet Temperature, avg = Air Inlet Temperature = 75.5~, Water-Main Pitot Tube Readings: and 164~F, avg = 163.5~F 156.0~F 77.5~, and 74.1~F, avg = 75.7~F varied widely (pulsating) from 3-1/2 to 8 in. 6

-ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - TABLE I (Concl.) Data on East Bay 3A R 3A L 3B R 3B L 3C R 3C L 3D R 3D L 3E R 3E L 4A R 4A L 4B R 4B L 4C R 4C L 4D R 4D L 4E R 4E L m-_ 1.85 1.92 1,74 1.76 1.67 1.82 1.45 1.65 1.50 1.73 1.48 1.50 AM." 1.90 1.94 1.80 1.75 1.76 1.82 * *Nm 1.47 1.60 1.50 1.70 1.50 1.50 1.875 1.93 1.77 1.755 1.715 1.82.. _ 1.46 1.625 1.50 1.715 1.49 1.50 418 485 546 676 523 500 523 761 637 534 492 400 464 472 677.5 520 672.5 523 507 506 59.8 60.0 60.1 61.0 60.4 60.0 60.1 60.0 60.0 59.9 60.0 60.0 6o. 0 60.0 62.5 59.6 60.7 60.0 60.i 420 485 545 665 520 500 521 759 639 534 492 400 464 453 681 520 664 523 506 506 thermocouple. East Bay Avg = 539.9 Cold-Junction Temperature (Cell No. 3) = 26.2~C = 79.2~F Cold-Junction Temperature (Cell No. 4) = 27.8~C = 82.0~F Water Inlet Temperature = 163~ and 162~F, avg = 162.5~F Water Outlet Temperature = 154.5~ and 154*5~F, avg = 154.5~F Air Inlet Temperature = 75.7~ and 68.6~F, avg = 72.1~F 7

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE II SUMMARY OF EXPERIMENTAL DATA OF EAST BAY DATA OF MONDAY, MAY 2, 1955 Arrangement of Tubes: three in line, fourth row staggered Time of test start: 2:00 P.M. Time of test finish: 2:15 P.M. Fan in Full-Pitch Position 3A 1.45 1.38 1.415 780 60 780 3B 1.40 1.42 1.410 743 59.8 745 3C 1.35 1.38 1.365 622 60.0 622 3D 1.30 1.28 1.290 806 60.1 804 3E 1.30 1.30 1.300 767 60.8 756 4A 1.47 1.46 1. 465 797 60.0 797 4B 1.18 1.18 1.180 857 60.0 857 4C 1.17 1.17 1.170 792 59.9 793 4D 1.25 1.26 1.255 891 60.2 889 4E 1.32 1.32 1.320 827 61.0 813 Overall Avg = 1.317 785.6 *Iron-constantan thermocouple Reading No. 1 2 5 3 4 Avg Cold Junction Temperature,~C Water Inlet, ~F Water Outlet, ~F Air Inlet Temperature, ~F Orifice Meter Motor Amps 29.2 153.0 144.5 81.0 29.0 153.5 143.5 81.5 66 inches 41.0 152.0 143.5 80.0 water 41.0 152.5 143.0 29.1~C 152.7~F 143.6~F 80.8 66 in. water 41.0 amps 41.0 (Barometric pressure 29.87 in. Hg, obtained from airport) 8

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE III SUMMARY OF EXPERIMENTAL DATA OF WEST BAY DATA OF MONDAY, MAY 2, 1955 Arrangement: triangular-pitch bay Time of test start: 2:'0 P.M. Time of test finish: 2:50 PoM. 1A 1.10 1.13 1. 115 548 62.4 526 lB 0.90 0.90 0. 900 761 60.0 761 1C 1.30 1. 31 1.305 613 60.0 613 1D 1.05 1.07 1.060 666 60.1 664 1E 1.05 1.10 1.075 686 60.4 681 2A 1.15 1. 11.140 554 60.0 554 2B 0.90 0.90 0.900 814 59.8 816 2C 0 0.9 0.96 0.940 535 60.2 532 2D 0.88 0.90 0.890 822 60.1 821 2E 1.02 1.00 1.010 650 60.0 650 Overall Avg = 1.0335 661.8 *Iron-constantan thermocouple Reading No. 1. i2 2 1 3 4 Ag Cold Junction Temperature, ~C Water Inlet, ~F Water Outlet, ~F Water Orifice Ambient-Air Thermometer Motor Amps 35.4 150 142 79.5 43 34.6 150 141 66 inches 79.5 43 35.1 147.5 140 of water 79.5 42 147.5 139 79.5 42 35.03~C 148.7~F 140.5~F 66 in. water 79.5~F 42. 5amps Wind velocity - 935 ft/min (Barometric pressure 29.87 in. Hg, obtained from airport) 9

7- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - Table III summarizes the field test data obtained on the new west bay of the jacket water cooler on Monday, May 2, 1955. This bay contained 178 tubes on a 2-/16-in. triangular pitch arrangement. It should be noted that the three tables also contain the overall average anemometer velocities in ft/min, average inlet and exit air temperatures, and water inlet and exit temperatures. A Pitot tube on the centerline of the 16-in. water main gave widely fluctuating readings of from 3-1/2 to 8 in. of meriman oil pressure on Thursday, April 28 (Table I). This Pitot tube was replaced by an orifice plate (1125 in. diameter) which gave reliable pressure-drop readings on Monday, May 2, (Tables II and III). A University of Michigan Pitot-static tube, five feet long, was taken along for making an air-side pressure-drop profile. The field units would not acc mmodate this Pitot tube. The units had a small hole which would accommodate a small-diameter tube (about 1/8 or 3/16 in.). Mr. Roberts had a makeshift arrangement whereby a copper tube was inserted through this hole for making a single-point pressure measurement. These data are contained under Section X of this report. B. PRELIMINARY ANALYSIS OF THE DATA 1. Fouling on the Water Side and Air Side. —The problem of determining a suitable water fouling factor for the water side in analyzing Thursdayts data was solved by using an experimentally determined fouling factor on a sample of a tube cut from the field unit. The experimentally determined value was found to be 0.0013, based on the liner area. The data for determining this value are given in Section VIII of this report. The selection of a proper fouling factor for the test data taken on Monday is open to question. The units tested were new units shipped from the Happy Co. of Tulsa, Oklahoma, for this purpose. There is a strong argument in favor of assuming no fouling, since the units had been on stream less than two days. In order to show the influence of some fouling on the computed results, these data were analyzed under the assumed conditions of zero fouling and fouling of 0.001 for the water side. No fouling was assumed for the air side. 2. Thursdayts Data.-The water-side flow-rate measurements obtaine& by means of a Pitot tube were unreliable due to rapid fluctuations of the manometer over a wide range (3-1/2 to 8 in.). An attempt was made to determine the performance of the unit, even though the fan pitch was known to have been varying during the test run. It was assumed that the air velocity and temperature profile (see Table I) would give an indication of the performance of the unit. The calculations are presented in Appendix A.

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - The results of the analysis are summarized below. Air face velocities (indicated): East Unit = 508 std. ft/min West Unit = 445 std. ft/min Average temperature rise of the air: East Unit = 65.7~F West Unit = 56.1~F Apparent-mass flow rate of air: East Unit = 825,000 lb/hr West Unit = 723,000 lb/hr Apparent heat duty: East Unit = 13,000,000 Btu/hr West Unit = 9,730,000 Btu/hr Total 22,730,000 Btu/hr The jacket water cooler was supposed to handle 21,492,000 Btu/hro It was quite evident that the above figure of 22,730,000 was in error, since it was a known fact that the unit was actually operating considerably below design load as specified by the specification sheet for the unit. This is further emphasized by the quantity of water that would have had to be flowing in order to transfer the indicated load. This calculation is also given in Appendix A and is summarized below: Required water flow rate: East Unit = 1,625,000 lb/hr West Unit = 1,390,000 lb/hr Total = 3,015,000 lb/hr Computed water flow rates: East Unit = 9.33 ft/sec West Unit = 8o00 ft/sec (The design water flow rate was approximately 5.5 ft/sec. ) On the basis of 3,015,000 lb/hr of water flowing through the tubes and a heat duty of 22,730,000 Btu/hr, an overall coefficient of 121 was obtained on the east unit, whereas a value of 82.8 was obtained on the west unit, based on the liner area. The water-side coefficients as computed were 2520 and 2230, respectively, based on the inside area. The corresponding ho values were 154.5 and 97o0, based on liner area. The water-film resistance is negligible in comparison with the air-side resistance. The fouling resistance used in the calculations was determined experimentally on a sample tube by measuring the overall performance of the tube before and after cleaning out the inside of the tube with steam plus a hot-water jet. The value determined was 0,0013 for inside fouling, based on liner area (see Section VIII of this report). 11

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The computed air-side coefficients are obviously incorrect, for if such coefficients were actually realized, the units should have performed considerably better than indicated, It is quite apparent that one or more inconsistencies in the data are influencing the computed results. A discussion of the inconsistencies of the Thursday and Monday data is given in Section II-C. The resolution of the inconsistencies is given in Section II-Do This is followed by an analysis of the revised data in Section II-E. 3. Monday's Data. - a. Three-in-line, fourth-row-staggered unit. The data obtained on this unit are given in Table II and are of such a nature that both air-side and water-side heat loads may be determined, The calculations are presented in Appendix B and are summarized below. Apparent air-face velocity = 747 std. ft/min Air-side heat duty = 13,900,000 Btu/hr Apparent air-side coefficient = 166,5 based on liner area (ri = 0) Water-side heat duty = 9,580,000 Btu/hr (assuming one-half the water passed through this unit) A comparison of the water-side heat duty with the air-side heat duty indicates a discrepancy of 4,520,000 Btu/hr. The air-side heat duty is 48% higher than the water-side heat duty. The design heat load for the unit is 21,492,000 Btu/hr for the two bays, or 10,746,000 Btu/hr per bay. The air-side coefficient of 166.5 (see Appendix B Section 6) was based on the air-side heat duty and therefore is questionable in view of the lower waterside heat duty. Another air-side coefficient can be computed, assuming the water-side heat duty is correct and that the measured air temperatures are correct. This assumes that the air velocity profile is incorrect. Such a calculation is presented in Appendix B Section 9. The computed necessary air velocity is 504 ft/min (face velocity) at standard conditions, and the corresponding air-side coefficient is 110, based on the liner area. The computed velocity of 747 ft/min based on air-side data, is 48% greater than the value of 504 computed from the water-side heat load. Also, the air-side coefficient of 166.5 is 51.5% higher than the corresponding value of 110 computed from the water-side heat duty. In all the above cases the waterside fouling was assumed to be zero. If the water-side fouling is assumed to be equal to 0.001, based on the water-side area, the ho computed from the water-side data is 126, based on the liner area. The outside coefficient, ho, based on the air-side test data, is 205, based on liner area. The air-coefficient value of 205 is 62.8% greater than the corresponding value of 126 (see Appendix B, Section 6). 12

F- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Since the water-side heat duty is simple to determine accurately with an orifice plate, and water-temperature drop is -easy to measure, there is every reason to believe that the lower air velocity and lower air coefficient computed on the basis of the water-side heat duty are representative of the air-side performance. The resolution of this point is presented in Section II-D of this report. b. Triangular-pitch unit. The data obtained on this unit are given in Table III and are of such a nature that both air-side and water-side heat loads may be determined. The calculations are presented in Appendix C and are summarized below, Apparent air-face velocity = 626 ft/min Air-side heat duty = 12,410,000 Btu/hr Apparent air-side coefficient = 174 based on liner area (ri = 0) Water-side heat duty = 8,450,000 Btu/hr (assuming one-half the water passed through this unit) A comparison of the water-side heat duty with the air-side heat duty indicates a discrepancy of 3,960,000 Btu/hr. The air-side heat duty is 47% higher than the water-side heat duty. An air-side coefficient can be computed on the basis of the water-side heat duty, assuming the measured air temperatures are correct. This assumes that the measured air velocities are incorrect. Such a calculation is given in Appendix C, Section 7, The air-side-face velocity so computed is 428 ft/min and the corresponding airside coefficient based on liner area is 117. The computed air-side velocity of 626 from the air velocity profile is 47% higher than the corresponding velocity of 428 computed from the water-side heat duty. The apparent airside coefficient of 174 is 49% higher than the corresponding coefficient of 117 computed from the water-side heat duty. In the above discussion the water-side fouling factor was assumed to be zero. The water-side heat duty appears more reliable than the air-side duty, and the lower air velocity and lower air coefficient give a more reliable indication of the performance of the unit. The discrepancy is resolved in Section II-D of this report. C. DISCUSSION OF INCONSISTENCIES The analysis of Thursday's data, presented in Section II-B of this report, indicated that there was reason to believe that the air-side heat duty was too high. Because of the poor method of measuring the water-main velocity, no indication of the discrepancy between the air-side and waterside heat loads could be made. 13

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - The analysis of Monday's data on the new three-in line, fourthrow-staggered unit indicated a discrepancy of 48% in heat duty between the air side and water side, The analysis of Mondayts data on the new staggeredpitch unit indicated a discrepancy of 47% between the air-side and waterside heat duties. It is therefore presumed that Thursday's air-side heat duty is about 48% greater than the water-side heat duty at the time of the run. The problem of resolving this dilemma resulted in considerable difficulty. A survey of the situation indicates that a number of possible measurements could have been made in error. These are as follows: 1. The exit air temperature profile was measured, using ironconstantan thermocouples and a Minneapolis-Honeywell potentiometer. 2. The inlet air was measured, using several ASTM distillation thermometers and an iron-constantan thermocouple. 3. The inlet and exit water temperatures were measured, using ASTM distillation thermometers in thermowells packed with grease. 4. The water flow rate was measured using a new sharp-edged orifice, having an ID of 11.25 in. placed in the 16-in., schedule-20-pipe water line. 5. The air velocities were measured, using a Taylor anemometer (in a six-in, stove-pipe duct) and a stop watch. The above items will be discussed in detail. The twelve thermocouples were made up at The University of Michigan and checked against a Bureau of Standards platinum -platinum rhoduim thermocouple and against Bureau of Standards glass thermometers, using a Leeds and Northrup precision potentiometer as well as the Minneapolis-Honeywell potentiometers used in field testing at Lovington, N. M. (before the field tests were made). Several of the thermocouples were rechecked after returning to the Universityo The potentiometer was also checked and found still to be operating satisfactorily. In addition to the above, the reference junction temperature was measured at Lovington with a calibrated Sargent thermometer. The air thermocouples were checked against ambient —air thermometers at the side of the tests and were found to be reading correctly. The temperatures of the air leaving the tube bank was measured, using the thermocouples and potentiometer described above. Stovepipe ducts, six in. in diameter and placed on top of the unit being tested, were used to direct the flow of the air from the tube bank past the thermocouples without mixing with ambient air. There exists the possibility that the readings of the thermocouples were high due to radiation from the stovepipes to the thermocouples. This possibility is discussed in more detail in Section II-D. 14

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - The inlet and outlet water temperatures were measured by means of ASTM distillation thermometers. This particular measurement is the most sensitive of the series. Small errors in the difference between the inlet and outlet temperatures make a large difference in the water-side heat duty. This particular type of thermometer was designed for use in laboratory distillation tests and in general is thought to be quite reliable to within + 0.25~F. During the test, the thermometer positions were occasionally reversed (outlet to inlet), with no apparent change in reading. Thus, the inlet and exit water temperatures are believed to be reliable~ A new sharp,-edged-orifice plate was installed in the outlet water line of the jacket water cooler during the "turnaround." Sharp-edged orifices are used extensively in industrial flow measurements and are capable of producing precise measurements of the quantity of flow through a pipe. Two possible sources of error can arise in the use of this device. The first can be an incorrect reading of the pressure drop occurring across the orifice. The second could be the use of a rounded-edge-orifice plate. The orifice plate was newly fabricated for this particular test and had sharp edges; the pressure-drop recording instrument was also new and had been installed for the test period. This instrument had been zeroed in on the morning of the test and should be considered reliable. The computed water velocity from the orifice pressure-drop measurements are therefore considered quite reliable. The air velocity profiles were made by using a Birams type vane anemometer manufactured by the Taylor Instrument Co. The instrument has a calibration card furnished by the company, the calibration in a wind tunnel having beenmade against a Bureau of Standards calibrated anemometer which had also been certified by the National Physical Laboratory of England. A possible source of error could arise as a result of damage of the instrument in transportation to Lovington, N.M. This possibility of error was eliminated by comparison of the performance of the instrument against a new instrument of the same type after its return to the University. The performance-check runs indicated that the instrument used in the field tests still was working properly. The stop watch used was also checked and found to be keeping correct time. It was therefore concluded that the instrument readings and corresponding time intervals as measured in the field were correct. In making the air velocity- and temperature-profile measurements it -was essential that the ambient-air wind flowing across the top of the table bank be kept away from the anemometer and the thermocouples. In order to accomplish this, stovepipe sections six in. n diameter and twelve in. long were obtained from a local hardware dealer. These sections were used as ducts to conduct the air'leaving the top row of tubes to the anemometer'and to the thermocouples. The ducts were long enough to permit mixing of the air 15

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN before reading the anemometers or the thermocouples. The possibility existed that the duct influenced the anemometer readings. An investigation of the influence of duct diameters on anemometer readings was undertaken after the University research group had taken up the problem with the Taylor Instrument Co. and found that the manufacturer of the instrument knew that there was such an influence factor but did not know the magnitude of the correction factor. The determination of this correction factor is discussed in Section II-D of this report. D. RESOLUTION OF INCONSISTENCIES 1. Reviewo-In the previous section of this report it was pointed out that the water-side data were apparently more reliable than the air-side data. Examination of the air-side data and test technique indicated that either the air-side temperatures were incorrect or the air-side velocities were incorrecto Two possible influence factors could have caused the inconsistent computed heat load obtained from the air-side data. These were (a) radiation affecting the thermocouple readings and (b) a duct factor affecting the anemometer readings. Either one or both of these factors could have affected the data. These factors will be discussed in detail. 2. Radiation Correction Factor.-During the summer a number of experiments were conducted on the roof of the East Engineering Building, using a 6-in.-stovepipe duct and one of the thermocouples used in the field tests. The first tests made were misleading, as it appeared that there might have been a sizable radiation factor. Later tests showed that the rise in temperatures attributed to the sun was actually a conduction heat transfer heating up stagnant air in the duct. In the early tests no forced draft was used. Later it was found that if the air was in turbulent flow past the duct and through the duct, this temperature rise largely disappeared. We were unable to determine exactly what the radiation correction factor amounted to, but it is known that under certain conditions such an error may be introduced. Since the correction factor appeared to be slight, it could never begin to account for the 50% discrepancy in heat loads. It was therefor concluded that the error was in the air velocity measurement. 3. Air Velocity Correction Factor.-The possibility of a large duct correction factor took on added importance since an air test apparatus for studying bond resistance of bimetal tubes was under investigation by the research group at the University. A major investigation of this factor was undertaken and the results indicated that a large correction factors isntroduced when four- and six-in. ducts are used with a four-in. Taylor anemometer. 16

I I- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - I This duct investigation resulted in Report No. 37 entitled "Investigation of the Performance of Vane-Type Anemometers in a Four-Inch Duct." This investigation was extended to cover the six-in. duct used in the field tests. The investigation involved the use of a water-calibrated air rotameter and critical-flow orifices for measuring the flow of air through the anemometer. The test arrangement and test procedure used in conducting the investigation are covered in Report No. 37 and will not be reproduced here. The results of the tests using the anemometer in a six-in. duct resulted in Fig. 2. The slope of the calibration line in Fig. 2 is 0.636. The actual air velocity flowing through the six-in. duct is 63.6% of that indicated by the anemometer when placed in the duct. The measured air velocities are therefore 57% higher than the actual velocities. This type of anemometer reads correctly, according to the manufacturer, when used in a duct six to eight times the diameter of the anemometer (24- to 32-in. duct for the test anemometer). Using the instrument in a smaller duct results in a change in velocity profile of the air moving past the vanes of the instrument, which in turn changes the performance of the instrument. This is due to the fact that this is an inertial type of instrument. The recomputation of the field test data is presented in Appendix D and is summarized in Section II-E of this report. E. REVISED ANALYSIS OF THE FIELD TEST DATA 1. General Considerations.- -The recalculations of the field test data, including the 0.63,6 duct correction factor on the air anemometer readings, are presented in Appendix D and are summarized below. 2. Thursdayts Data.. — a. East bay. Reference is made to Appendix A where the indicated standard air velocity is 508 ft/min. Introducing the duct correction factor of 0.636 reduces this velocity to 323 ft/min. This reduces the indicated air-side heat load of 13,000,000 Btu/hr to a corrected value of 8,290,000 Btu/hr. The corresponding water velocity computed from the air-side data is 5.95 ft/sec. The revised water coefficient becomes 1766. The corrected overall coefficient based on liner area is 77 as compared to the uncorrected value of 121, The revised air-side coefficient is 90.8, based on liner area, as compared with the earlier value of 154.5 b. West bay. Reference is made to Appendix A where the indicated standard air velocity is 445 ft/min. Introducing the duct correction factor of 0.636

i - ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN - _1 I) ~r4) CM tNIio oI9 NI Aiio3A v niov NIN / I/.L= a-LS'.LOng,,9 NI A.IOO'13A I 11V "vn..ov Fig 2. Calibration of a four-inch vane-type anemometer in a six-inch-diameter duct.

I - ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN reduces this velocity to 285 ft/min. This reduces the indicated air-side heat load of 9,730,000 Btu/hr to 6,180,000 Btu/hr. The corresponding water velocity computed from the revised air velocity is 5.09 ft/sec. The revised water coefficient becomes 1552. The corrected overall coefficient based on liner area is 52.7 as compared to the earlier value of 82.8. The revised air-side coefficient is 59.1, based on liner area, as compared with the earlier value of 977.7 c. Average of east bay and west bay. The computed results given above are compared and averaged below. East West Avg Face Velocity 323 283 303 Heat Duty 8,290,000 6,i80,000 7,235,000 Uo 77.0 52.7 64.9 ho 90.8 59.1 75 d. Discussion. The average air-face velocity of 303 ft/min is about one-half the design-face velocity of 600 ft/min. This velocity is believed to be on the edge of the transition region between laminar and turbulent flow. Inspection of the data for this unit, given in Table I, indicates that large areas of the two bays have velocities significantly below the average values. It is possible therefore that the air flowing past the tubes in these regions of low velocity would not have a fully establishdThe a fully established turbulencirfilm coefficients for these areas would be much lower than that predicted by any correlation obtained from turbulent air-film data and extrapolated back to this velocity. 3. Mondayts Data on Three-Rows-in-Line, Fourth-Row-Staggered Unit, Assuming a Water-Side Fouling Factor of 0.001. —Reference is made to Appendix B where the average standard air-face velocity was originally computed to be 747 ft/min. Introducing the duct correction factor of 0.636 gives 475 ft/min. This results in a new air-side heat duty of 8,840,000 Btu/hr (see Appendix D). The corrected water velocity turns Out to be 5.58 ft/sec. The revised overall coefficient is 97, based on iner area, and the recomputed air-side coefficient is 11845^ based on the liner area. A comparison may be made between the revised computed performance given above and the corresponding information computed from the water-side heat load. The water-side heat duty (see Appendix B) was computed to be 9,380,000 Btu/hr as compared with the revised value from the air-side data of 8,840,000 Btu/hr. The air-side heat duty is only 5.76% below the waterside heat duty. The check is reasonably close. In Appendix B, a calculation 1 - 19

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN of the air-side coefficient is made, based on the water-side heat load and the air temperatures, assuming the air velocity was incorrect (later proved by laboratory test). The computed air velocity turned out to be 504 ft/min as compared to the revised air velocity of 475 ft/min. The computed Uo based on liner area is 103 as compared to the revised air velocity Uo value of 97. The corresponding ho based on liner area was 126 as compared to the revised air-side ho value of 118.5. It is quite apparent that the introduction of the anemometer duct correction factor brings the air-side performance on this unit into reasonable agreement with the water-side predicted performance. 4. Mondayts Data on the Triangular-Pitch Unit, Assuming a WaterSide Fouling Factor of 0.001. —Reference is made to Appendix D, Section B-2, where the air-side data are recomputed, using the duct correction factor of 0.636. The uncorrected air velocity (std) was 626 ft/min, whereas the corrected value becomes 398 ft/min. The original computed air-side heat load of 12,410,000 Btu/hr is reduced 7,900,000 Btu/hr. The uncorrected overall coefficient of 159 is reduced to 101.3. The uncorrected air-side coefficient, based on liner area as computed in Appendix C, was 216. The corrected value as given in Appendix D, Section B-2,is 125, based on liner area. In Appendix C a calculation of the air-side performance, using the water-side heat load and inlet and exit air temperatures, is presented. The calculations indicate a computed air-side velocity of 428 std ft/min, an overall coefficient of 108, and an air-side coefficient of 135 based on liner area. These values compare favorably with the corresponding corrected airside values of V = 398 ft/min, Uo = 101.3, and ho = 125. It is interesting to compare the revised air-side heat load of 7,900,000 Btu/hr with the water-side heat load of 8,450,000 Btu/hr. The air-side heat load is only 6.5% below the water-side heat load. It is quite obvious that the introduction of the anemometer duct correction factor brings the air-side performance on this unit also into reasonable agreement with the water-side predicted performances The discrepancy between the air-side and water-side heat loads for the three-in.line, fourth-row-staggered unit amounted to 5.76% in the same direction, as compared to 6.5% for the triangular-pitch unit for an average discrepancy of only 6.1%. 5. Mondayts Data Re-evaluated, Assuming No Fouling on the Water Side.- a. Three-in-line,. fourth-row-staggered unit. The air-side coefficient recomputed from the air-side data, assuming 20

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN no fouling on the water side, is 105, based on liner area, or 7.64, based on the outside area (see Appendix D). These values compare with 118.5 and 8.65, with an assumed water-side fouling factor of 0.001. The assumption of no water-side fouling results in a decrease of 12.8% in the computed airside coefficients. This emphasizes the importance of water-side fouling in analyzing the data. b. Triangular-pitch unit. The air-side coefficient recomputed from the air-side data,.assuming no fouling on the water side, is 110, based on the liner area, or 8.02, based on the outside area (see Appendix D). These values compare with 125 and 9.13, with an assumed fouling factor of 0.001 on the water side. The assumption of no water-side fouling results in a decrease of 13.6% in the computed air-side coefficients. c. Discussion. Examination of the above results indicates a significant difference in the computed performance of the units when the inside fouling factor is assumed to be zero or is assumed to be 0.001. Fouling rates vary exponentially as a function of time. Therefore, there probably is a stronger argument in favor of the case in which no fouling is assumed, since in all probability several months presumably would have to pass before a degree of fouling equal to 0.001 would be developed. 6, Summary.-Fol comparison purposes the final computed performance of the units are summarizedc below: TABLE IV TABULATED RESULTS Thursday s Data East bay V (air-face velocity) = 323 std ft/min Q (air side) = 8,290,000 Btu/hr Uo (liner area) = 77.0 (ho)L (liner area) 90.8 ho (outside area) = 6.63 (At)LM = 47..8~F West bay V (air-face velocity) = 283 std ft/min Q (air side) = 6,180,000 Btu/hr 21

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN UO (liner area) (ho)L (liner area) ho (outside area) (At)LM 52.7 59.1 4.32 52.3~F Average Performance of Unit on Thursday, April 28, 1955 V (air-face velocity) Q (air side) Uo (liner area) (ho)L (liner area) ho (outside area) (At)LM 303 std ft/min 7,235,000 Btu/hr = 64.9 - 75.0 5.48 50.05~F Monday's Data, Three-Rows-in-Line, Fourth Row Staggered Bay V (face velocity) air side V (face velocity) water side V (face velocity) average Q (air side) Q (water side) Q (average) Uo (air side, liner) Uo (water side, liner) Uo (average, liner) (At)LM 475 std ft/min 504 std ft/min 490 std ft/min 8,840,000 Btu/hr 9,380,000 Btu/hr 9,110,000 Btu/hr 97.0 103 100.0 = 40.5~F Assuming an inside fouling factor of zero: (ho)L (liner) air side (ho)L (liner) water side (ho)L (average) ho (outside) air side ho (outside) water side ho (outside) average 105 110 107.5 7.65 8.05 7.85 Assuming an inside fouling factor of 0.001: (ho)L (liner) air side (ho)L (liner) water side (ho)L (liner) average ho (outside) air side ho (outside) water side ho (outside) average 118.5 126 122.3 8.65 = 9.18 8.91 Monday's Data, Triangular-Pitch Unit V (face velocity) air side V (face velocity) water side 398 std 428 std ft/min ft/min 22

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - V (face velocity) average = 413 std ft/min Q (air side) = 7,900,000 Btu/hr Q (water side) = 8,450,000 Btu/hr Q (average) = 8,175,000 Btu/hr Uo (air side, liner) = 101.3 Uo (water side, liner) = 108 UO (average, liner) = 104.7 (At)L5 -= 34.7~F Assuming an inside fouling factor of zero: (ho)L (liner) air side = 109.8 (ho)L (liner) water side = 117 (ho)L (liner) average = 113.4 ho (outside) air side = 8.02 ho (outside) water side = 8.56 ho (outside) average 8.29 Assuming an inside fouling factor of 0 001: (ho)L (liner) air side (ho)L (liner) water side (ho)L (liner) average ho (outside) air side ho (outside) water side ho (outside) average 125 135 130 = 9.13 9.85 9.49 23

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN III. HEAT TRANSFER CORRELATIONS FOR TUBE BANKS A. CORRELATIONS FOR AIR-SIDE COEFFICIENTS FOR THREE-IN-LINE, FOURTH-ROWSTAGGERED UNITS 1. 651-A-B-C. —The Wolverine Tube Division has published a series of threebrochures entitled "Data Sheet 651-A," published July 1, 1950, "Data Sheet 651-B," published July 1, 1950, and "Data Sheet 651-C," published July 1, 1950, and revised February 1, 1951. This series of data is not reproduced in this report. The data as published are presented in tabular form, giving the length of various-sized finned tubes having various numbers of fins per inch required to transfer 10,000 Btu/hr (condensing steam) as a function of the maximum air velocity and the log mean At driving force. The tabulated data are of such a nature that one can readily compute the air-side coefficients used in making up the tables. A sample calculation is presented in Appendix E. Figures 3, 4, and 5 give the ho-vs-Vmax curves for 651 A, B, and C, respectively. Figure 6 presents a single line representing all the three curves from Figs. 4, 5, and 6, which has the following equation: ho = O.2O(Vmax)' where ho = air-side coefficient based on outside area and Vmax = maximum air velocity at the minimum cross section, ft/min. This equation can be converted to a corresponding relationship for a9-fins-pez inch tube, giving the air-side coefficient, ho, based on the liner area, as a function of face velocity. The relationship is computed as follows: (ho)liner = 13.7(ho)outside area Vmax = 2.428 Vfacel 24

I LL 0 c0 I I n" 1 Legend Vmax /Vface * 5 Fins/inch 1.80 a 7 Fins/inch 1.90 lo/ t DF 1 ] 1 t - ~o II Fins/inch 1.952 I L X_______ I_ _____ ____ __ _I I m z rt Z m m z 70 m Il mC v7 rr m0 <: rz m 70 -< 0 In poo Cn Z 10 100 1000 VMAX-FT/MIN 10o Fig. 3. Wolverine tube data sheet 651A, published July 1, 1950. Outside coefficient, calculated from data of tables 2 and 4, 6 and 8, 10 and 12, and 14 and 16.

IC FLL 0 CM - I l. ro o\ DO -— __ ^_ __ -_ 7 I C _- - -7 —---- X Legend: Vm ax/Vface *5 Fins/inch 2.00 / 7 Fins/inch 2.00 o 9 Fins/inch 2.087 100, 13|~~~~~~~~~ 11 Fins/inch 2.076. in s/incUU_.00U m z AC) z m m 73 z 0 73 m m > 73 rn I z t~ - -= m z < 03 m 0 I Z AA IU IUU100 IUUU IUooo VMAX- FT/MIN Fig. 4. Wolverine tube data sheet 651B, published July 1, 1950. Outside coefficient, calculated from data of tables 2 and 4, 6 and 8, 10 and 12, and 14 and 16. i

F - Ic: m I o a Legend Vmax/Vface, ___ ___________ _ ___ * ___5 Fins/inch 2.28 ^// A 7 Fins/inch 2.34,,0 o 9 Fins/inch 2.428 o I Fins/inch 2.416,- r _____ ^_____ ___ __ __ _ ~______________ ___ __ __ ~ ~______________ ___ __ _ ~,.j/n z m z m m m m m Zs c~ ma *, ro Z z m 0 -'n oo I z I0 IUU IUUU VMAX-FT/MIN Fig. 5. Wolverine tube data sheet 651C, published July 1, 1950. Outside coefficient, calculated from data of tables 2 and 4, 6 and 8, 10 and 12, and 14 and 16. Iu,)u

u c a 0 F a.9 IUU 651-A - 651-B- j - 651 ---- ho =020(Vmoxp' r10 0 o II Liner diameters'__________ ___ __-__ _ _____ - - - - _ ___ 651-A: 3/8" OD 651-B: 5/8"OD ^^ ] ] 1 / 1 < X -L 4 tk0 ~ 651 -C: I"OD m z Z GI m m -r m rm 3a Z m o -r 0 zn C3 Z 10 100 1000 VMA- FT/MIN Fig. 6. Wolverine tube data sheets 651A, 651B, and 651C, published July 1, 1950. Outside coefficients, calculated from data of tables 2 and 4, 6 and 8, 10 and 12, and 14 and 16. 10,0o

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Substituting, (ho)liner = (13.7)(0.20)(2.428 Vface) 0 0.6 (ho)liner = 467(Vface) The air-side heat transfer coefficients were computed from the tables indicated by assuming a steam-condensing coefficient of 1000 with no allowance for fouling, metal, or bond resistance. The resistance of the metal and of the fouling factors for steam condensing to air are slight in comparison with the resistances of the air-side film coefficient. The effect of bond resistance is unknown. Examination of Figs. 35 4,.5 and 6 indicates that a single correlating line was probably used by someone in the preparation of the tables. The correlation report, Report No. 30, presents an equation (on page 6 of the report) which indicates the influence of the number of fins per inch on the air-side coefficient. This equation indicates that the coefficient varies inversely with the square root of the number of fins. No such factor was used in 651A, B, or CJ Figures 7, 8, and 9 graphically present the pressure-drop information for the air side of the tubes as given in 651A, B, and C. Figure 10 summarizes the curves presented in Figs. 7, 8, and 9. In Fig. 10 the air-side pressure-drop equation for a 2-in.,-OD tube having 9 fins per inch and a 1-in. liner is AP = 1.84 x lO'6 (Vface)89 where AP = inches of water for four rows deep (three in. line, fourth row staggered) and Vface = face velocity, ft/min. a. Prediction of coefficient and pressure drop, using 651, Thursday's test data. Reference is made to Table IV in which the average face velocity is reported as 303 ft/min with an ho on liner area of 75.0. Substituting this face velocity into the 651A, B, and C equation gives I - I - 29

- ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN I ULL O I o I <3 10 -- ----- ~I~~~ W.O Legend Vmax/ face * 5 Fins/inch 1.80,/ ___________ ______a 7 Fins/inch 1.90 o 9 Fins/inch 2.00 O II Fins/inch 1.932 0.1 ______ J 100 1000 Vface -FT/ MIN 10000 Fig. 7. Wolverine tube data sheet 651A, published July 1, 1950. Air-side pressure drop, calculated from data of tables 5, 9, 13, and 17. 30

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - 0 I LL 0 O 2 0. 10 I.0 Legend max Vface o 5 Fins/inch 2.00 _ 7 Fins/inch 2.00 o 9 Fins/inch 2.087 13 I Fins/inch 2.076 0.1 - o.L I 1 9_ 1 _ I_ I 1 _ 1 1 111 111 J 100 1000 Vfce- FT/MIN 10000 Fig. 8. Wolverine tube data sheet 651B, published July 1, 1950. Air-side pressure drop, calculated from data of tables 5, 9, 13, and 17. 31

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 10r 1 I I I I I I I I I 1 I I I I 1 1 -!~~~~~~~~~~~~~~~~~~~~~~~~~~~. ===- _: == 11 ~~~~/ I ACXT0XI I,. == - /.. ===_ ______== = _ —-___.===__ _ L 1 I I I I IS II H /1 /I A/6i. I I- - I IL 0 IC. n lLJ z I Q. <3 I K Legend I______ I_,_ _I 71L___, * 5 Fins/inch a 7 Fins/inch o 9 Fins/inch 1 II Fins/inch I 2.28 2.34 2.428 2.416 I Vmax /Vface i / 0.1'100 1000 Vface - FT/MIN.A_ I 10000 Fig. 9. Wolverine tube data sheet 651C, published July 1, 1950. Air-side pressure drop, calculated from data of tables 5, 9, 13, and 17. 32

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - IO 10 ------- 651 - B AP= 1.60X 10/ (Vfoce)1"87 ~I ""'/ "\651-C AP=1.84 X 10-6'r^~~~~~ /// ~~~~~~~~~~(Vface)89 LL / // =(IIII )n z <a -,'/ Vf-ceFT/MIN Fig. 10. Wolverine tube data sheets 651A, 651B, and 651C. Air-side pressure drop, calculated from data of tables 5, 9, 13, and 17. 33

F- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1 (ho)liner = 4.67(Vface)06 = 4.67(305)~'6 145 The 651 value of 145 is 935 greater than the average value obtained from the field test data. The air-side pressure drop predicted by 651C is AP = 1.84 x 106 (Vface)'9 in. of water = 1.84 x 10-6 (303)1'89 = 0.09 in. of water. This value will be discussed in Section VI. b. Prediction of coefficient and pressure drop, using 651, Monday's test data for 3-in-line, fourth-row-staggered unit. In Table IV the air-side face velocity is given as 475 ft/min with an ho value of 105 (ri = 0) based on the liner area. The 651 equation gives (ho)liner = 4.67(475)0.6 = 188. The 651 equation gives a value that is 79% greater than the field-test-data value. The air-side pressure drop predicted by 651C is AP 1.84 x 10'6 (475)1e89 0.202 in. of water. This value will be discussed in Section VI of this reporto c. Prediction of coefficient and pressure for triangular-pitch unit, using 651, Monday's test data. The 651A, B,.and C air-side-coefficient equation and air-side pressure-drop equation are limited to three-in-line, fourth-row-staggered units and therefore are not applicable to this test data. I [

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 2. American Locomotive Co. Data. —The only test data in the University Research Project 1592 files bearing on the performance of threein.line, fourth-row-staggered units were supplied by the American Locomotive Co. under the date of September 15, 1949. Figure 11 presents the ALCO data for l-in.-liner, 2-in.-OD, 7-fins-per-inch Wolverine tubes. The figure presents ho, based on the outside area, as a function of face velocity and the pressure drop in inches of water for the four-row-deep bank, also as a function of face velocity. Unfortunately, the line which correlates the test data is for a 7-fins-per-inch tube. Equation 6, page 6, of the correlation report (Report No. 30) indicates that the air-side coefficient varies inversely with the square root of the number of fins per inch. Another line, corrected for the difference between 7 fins per inch and 9 fins per inch, using this equation, is presented in Fig. 11. Also indicated is the 651C correlating line. It is evident that the 651C line gives considerably higher values than indicated by the ALCO data. The equation of the ALCO test-data line is ho = 0.20 Vface06 (for a 7-fins-per-inch tube), where ho = air-side coefficient based on outside area. The corresponding equation in terms of liner area is (ho)liner = 2.16 ace for a 7-fins-per-inch tube). The equation of the ALCO line converted over to 9 fins per inch is (ho)liner = 2.42 Vface0 (for a 9-fins-per-inch tube). In Section II of this report the equation of the 651 coefficient line was given as (ho)liner = 4.67 Vface Taking the ratio of the ALCO equation for a 9-fins-per-inch tube to the equation of 651 gives ALCO(ho)liner 2.42 0 651(ho)liner 4.67 The ALCO line therefore gives a value which is 52% of that predicted by 651. The pressure-drop curve of ALCO, as indicated in Fig. 11, gives excellent agreement with the pressure drop predicted by the 65lCTline on the same plot. 35

I l I I I I I I I I I I I I I I I I I I I L I I I I I I I I I I I I I 1 t I I I I L I I I I I I I I 100 LL IV.^ / 0 3. it 0 twO Io LL 0 0 lL z a0 ON 3 U) w aI 5.0! LK1' ________// ---- - __i 7 Fins/inch-= -' 651-C P7-Toble9' 3C / / - LL 1.0 ~ I0 2 —-- -___ _ —--- ____ _. —2,/ 65i-C — ~'I ) 1,0- 9 Fins/inchGo IL 0 10 C Li C I0.3> 3 - ~~~~~~~~~~~~~~~~~~~~~~~~~0.' _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - - - - - _ __1 O.I ----------- -- --- -- -- I- - - --------------- --- -- -- - - - - ----- --- ----- --- -- -- - - _ _l~~~~~~~~~~~~~~~~t m z - z m Z 3 CS m r m: n! I -r I Z o *.I i -,<.rl 0'I Z. r z 100 200 300 400 500 1000 I 500 FACE VELOCITY FT/MIN 1000 1500 3000 10000 Fig. 11. ALCO Products Division, American Locomotive Company, heat transfer and pressure drop, Wolverine bi-metal x-fin tubes, 1" OD liner (7 fins) 3 in line, 4th staggered, September 15, 1949. I

L. ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN a. Thursday's test data, using ALCO equation. Reference again is made to Table IV in which the average of Thursday's data gave a face velocity of 303 ft/min and an ho of 75, based on the liner area. The ALCO line converted to 9 fins per inch gives the following ho on liner area for this velocity: 0.6 (ho)liner = 2.42 (Vface) 2.42(303)~' 75. The corresponding air-side pressure drop, as predicted by the ALCO curve of Fig. 11 for 7 fins per inch, is 0.115 in. of water as compared to 0.09 in. of water by 651C. b. Monday's test data, using ALCO equation for three-in-line, fourthrow-staggered unit. In Table IV the air-side-face velocity is given as 475 ft/min with an ho value of 105 (ri = 0), based on the liner area. The ALCO equation predicts (ho)liner = 2,42 (475)06 = 97.5. The ALCO pressure-drop curve, Fig. 11, predicts a value of 0.248 in. of water as compared to 0.202 in. by 651C. c. Monday's test data, using ALCO data for triangular-pitch unit. The ALCO information is strictly limited to three-in-line, fourthrow-staggered units and is not applicable to this unit. 5. Happy Co. Rating Curve.-The Happy Co. has furnished a rating sheet which is given in Fig. 12. The figure presents the outside coefficient ho, based on the liner area,as a function of the face velocity in std ft/min. The figure also presents the pressure drop in inches of water per row as a function of the same velocity. The data are presented for a 2-in.-OD tube having a 1-in. liner and 9 fins per inch. a. Thursday's test data, using the Happy Co. rating curves. Table IV gives an average face velocity of 303 ft/min and an average ho of 75, based on the liner area. The Happy Coo heat transfer curve, Fig. 12, predicts an ho of 130, based on the liner area: for this face velocity. The Happy Co. pressure-drop curve predicts a pressure drop of 0.024 in. of 37

ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN — 7 30C 0 LL ( 20C bJ LJ D z IOC o 9C L 8C cn < 7C tm ) L I" OD Wolverine trufin 1/2" High fins X 0.019" thick 9 Fins/in. ratio = 13.78 ft /ft2 i.I,C.C.cl ) ) I / r 2/i I I l I/I I1 / 2: CO) IJ I 0 I: LlJ )8 a. )7 a 0 I5. )6 c LlJ )5 er 03 n )4 w'L c) 003 )3 II 12 )2 < LLL a 0 o Q: I CD Li 0 IJ o 0 5Cv. } t'.C 50 1 —\.C 40 ----.0 30.0 20.0 0 / ------- ) I 100 200 FACE VELOCITY 300 400 FPM AT 70 F 600 800 1000 AND OFT ELEV. Fig. 12. The Happy Company heat transfer and pressuredrop rating curves. 38

-- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN water per row. For four rows deep the pressure drop would be 0.096 in. of water. b. Monday's test data, using the Happy Co. rating curve. In Table IV the air-side-face velocity is given as 475 ft/min with a corresponding.' ho value of 105 (ri = 0), based on the liner area for the three-in-line, fourth-row-staggered unit. The Happy Co. curve, Fig. 12, predicts a corresponding ho of 167 for this face velocity. This figure also predicts an air-side pressure drop of 0.055 in. of water per row. For a four-row-deep tube bank this amounts to a pressure drop of 0.22 in. across the bank. 4. Sunamary.-The field-test-data results are compared with the predicted values from 651C, ALCO dataf and Happy Co. rating curve in Table V. TABLE V COMPARISON OF PREDICTED COEFFICIENTS WITH TEST RESULTS FOR THREE-INE-LINE, FOURTH-ROW-STAGGERED UNIT Item (ho) liner AP in. 20 Thursdayts data computed (Vf = 303) 75 651C 145 0.09 ALCO (AP, 7 fins per inch) 75 0.115 Happy Co. rating curve 130 0.096 Monday's data computed (Vf = 475) 105 651C 188 0.202 ALCO (AP, 7 fins per inch) 97.5 0.248 Happy Co. rating curve 167 0.22 An examination of Table V indicates that the field-test coefficients agree reasonably well with the ALCO line converted for a 9-fins-per-inch tube and do not agree with the 651C predictions nor with the Happy Co. ratings. B. TRIANGULAR-PITCH CORRELATIONS 1. Correlation Report.. —The correlation report prepared by the University of Michigan Engineering Research Institute Project 1592, entitled "Correlation of Heat Transfer and Pressure Drop for Air Flowing Across Banks of Finned Tubes,"i Report No. 30, presents correlations that are restricted to triangular pitch arrangements.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Figure 13 presents the correlation for air coefficients as given on page 39 of Report No. 30. The correlation presents the correlating group as a function of the maximum air velocity in ft/min. The equation of the correlating line is Dr 0 56 ho = 19sW V 0 x 6 where ho = air-side coefficient based on outside area, Dr = root diameter, inches, Do = fin diameter, inches, N = number of fins per inch, S = tube pitch, inches, and Vmax = maximum air velocity, ft/min. A generalized heat transfer correlation for any gas as a function of the Reynolds number and the physical properties of the gas is also presented in Report No. 30, page 41. This correlation is reproduced here as Fig. 14. A third air-coefficient correlation is presented on page 45 of Report No. 30. This correlation gives the air coefficient as a function of the tube pitch, S, the number of fins per inch, N, and the root diameter, Dr and is reproduced here as Fig. 15. The correlation report also presents several pressure-drop correlations for air flowing across staggered-pitch tube banks. A correlation presenting air-pressure drop as the product (AP)De against maximum air velocity in ft/min with a parameter of K = NDo D 0.2 is given on page 42 of Report No. 30. The correlation is reproduced here as Fig. 16. An empirical pressure-drop correlation for air is also given on page 46 of Report No. 30. The figure is reproduced here as Fig. 17. A generalized pressure-drop correlation is presented on page 47 of Report No. 30 and is reproduced here as Fig. 18. The field test data taken on Monday on the triangular-pitch unit can be readily analyzed by the correlations of Report No. 30. Table IV gives the air-side-face velocity as 398 ft/min. This must be converted to a Vmax

- - - 200 0 150'o~):)c 0~Q^^9~ 0 I00 ~~~100 9~~0 00 - o o~F ~o ~,. o 0 o 80 o_ - o --- 80 I0 c } 0 ~ ~ 0 ~ 50 Q _ __ _- 0 _ _ _70 -0 - 0H 0 00 50 __ —- - ------- - 40,,_ 30 0....'______ _____ ___________ - 150 200 300 400 500 600 700 800 900 1000 1500 2000 3000 400C MAXIMUM AIR VELOCITY, FPM m z m z m m 50 I n 7m tS m 20 z m Z., m 73 — n Fig. 13. Heat transfer correlation for air.

10 IL (0 6 - l1 ~' ro R) m z CS z m m 73 z 50 m rt I m > 70 n z z - m O z n~ 1500 1 1 1 I _ L 1 I I I I I 100,000 200,000 300,000 600,000 800,00 1,000,000 /De GMAx. 2,000,000 Fig 14. Generalized heat transfer correlation.

1 90 70 0__________ _____ 0___ 0 8~~~~~~~~~~~~~0~~~~0 60 --- - --- -- __ ________3 0 _ Q C 70 0 l O' 5( 0 0 co )- 0.. c~~~~g, o o 0 0o 30i~~~c 1~~~~~00 0.o 1 I I I I o3 I ~~~o a ) DjO 0: o20 _'~ 00 _ 15 1,,, 200 250 300 400 500 600 800 1000 1500 2000 3000 5000 MAXIMUM AIR VELOCITY, FPM Fig. 15. Heat transfer correlation for air as a function pf 8, N, and Dr. m Z m z m m rn 70 z m frn z C:'1 r 7Z u( 0 iS o1 i

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN 10 8 6 4 3 LL 2 1.0 0.6 a. w 0.6 a. 0.4 I 0 0!! E_1 I T x 7 ~ l 1I lII.0 7 1 0.3 0.2 0.10 h n / p IV u %.F.%Opo I. I I.- -. 1 4 )51 j (-Vkj 0.06 I p K_ NDo 02 Dr -7! AtV.~ V-Vm# 200 300 400. 600 800 1000 2000 3000 4000 6000 _ _ _ I0,000 MAXIMUM AIR VELOCITY, FPM Fig. 16. Pressure-drop correlation for air. 44

I- ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN r I 0 o d a 10 a z _ a J 300 400 600 Ipoo 2P0o 3po 0ooo lo0ooo MAXIMUM AIR VELOCITY, FPM Fig. 17. Empirical pressure-drop correlation for air. 45

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 0 0 x 0 0 0 0 00 0 oo o o 0 0 0 O0 0 0 0 /@o 0o __ __ __ _0__ _o 50~ 0 0 00 V Q 00 0 0 ___o i~p.0oNU 0 0 000 0 0 Y0 - cO __ O o 0 0 0 LO I4 0 ^ oo od 0 -- ^ Oc ro C'XI (N K x A. I

r - ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - I and is equal to Vmax = 2.428 Vface = (2.428)(598) = 968. This value is used for computing the air-side coefficients and pressure drop by Figs. 13, 14, 15, 16, and 17, respectively. The calculations are presented in Appendix F and are summarized below. The ho computed by use of Fig. 13 is 9.9, by Fig. 14 is 10.6 and by Fig. 15 is 8.5, all based on the outside area. The corresponding values based on liner area are 135.5> 145.0, and 116.4, respectively. The air-side pressure drop by Fig. 16 is computed to be 0.356 in. of water. 2. Schmidt' s Correlation. —Mr. T. E. Schmidt published a correlation in an article entitled "Heat Transmission and Pressure Drop in Banks of Finned Tubes and in Laminated Coolers." The article appeared in the Institute of Mechanical Engineering and ASME Proceedings of the General Discussion on Heat Transfer, Section II, 186, London (1951). T. E. Schmidt' s heat transfer correlation applied to air is given on page 44 of Report No. 30 and is reproduced here as Fig. 19. The correlation can readily be used for predicting the air-side coefficient for staggered tube banks. The calculation of the air-side coefficient, using this correlation, is presented in Appendix G for the air-face velocity of 398 ft/min obtained in Monday's test on the staggered-pitch unit. The computed air-side coefficient is 7.73, based on the outside area, or 106, based on liner area. 3. Summar.y.-The analysis of the field test data is compared with the available correlations in Table VI. TABLE VI SUMMARY OF TRIANGULAR-PITCH ANALYSIS FROM AIR-SIDE DATA J (ho)liner ho AP Field test data (Table IV) (ri = 0) 110 8.02 Field test data (Table IV) (ri = 0,001) 125 9.1'3 Predicted by Fig. 13 136 9.9 Predicted by Fig. 14 145 10.6 Predicted by Fig. 15 6 8.5 -- Predicted by Fig. 19 (Schmidt) 106 7.73 -- Predicted by Fig. 16 --- 0.356.,.............._,_........__.,..._........ An analysis of Table VI indicates that the field test data check the Schmidt correlation. The field test coefficient is 19% lower, than that predicted by the correlation report curve for air (Fig. 13). 47

l 40 30 to d QO 0 z I 0 -C 20 10 9 m Z Z > m m )y 70 m m n 3000 4000 I z 3A C: m - — I 73 Z/~ - ------ n _ _____ 8 7 6 200 300 400 600 800 1000 2000 MAXIMUM AIR VELOCITY, FPM Fig. 19. Heat transfer correlation of T. E. Schmidt applied to air.

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN IV. PERFORMANCE OF THE THREE-IN-LINE, FOURTH-ROW-STAGGERED UNIT COMPARED WITH THE PERFORMANCE OF THE TRIANGULAR-PITCH UNIT Table VII presents a comparison of the performance of the two units (Monday's test data). TABLE VII COMPARISON OF TRIANGULAR PITCH AND THREE IN LINE, FOURTH ROW STAGGERED Three in Line, Triangular-Pitch UnitFourth Row Staggered _ _ i,~Fourth Row.Staggered Air Side Water Side Avg Air Side Water Side Avg Vface 398 428 413 475 504 490 Vmax 969 1039 1004 1154 1222 1188 Uo 7.4 7.9 7065 7.08 7.52 7.30 Uo(litner) 101.3 108 104.7 97 103 100 Qmillion 7.9 8.45 8.175 8.84 9.38 9.11 (At)mean 34.7 34.7 54.7 40.5 40.5 40.5 ho(ri = 0) 8.02 8.56 8.29 7.65 8.05 7.85 ho(liner)(ri = 0) 110 117 113.4 105 110 107.5 ho(ri =.001) 9.13 9.85 9.49 8.65 9.18 8.91 ho(liner)(ri =.001)125 135 150 118.5 126 122.3 A number of comparisons can be made between the performances of the two units. A ratio of the performance variables is as follows: Vface 3-4 Vface A Vmax 3-4 Vmax A 490 - 413 1.185 or 18.5% greater Uo 3-4 2. U0 "A = 765- = 0.954 or 4.6% less 7.65 49

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN &3. 3-4 9,110,000 = 1115 or 11.5% greater Q A 8,175,000 4. (At) 3-4 40o5 1.166 or 16.6% greater (At) A 334,7 5 (hA - r = 0. 948 or 5.2% less hoA ri= 0 8'9 The 18.5% greater air velocity flowing through the 3-4 unit resulted in a 4.6% lower overall coefficient Uo. The greater throughput of air in the 3-4 unit permitted the air to leave at a lower exit air temperature, which in turn resulted in a 16.6% larger At driving force for the 3-4 unit. This greater At driving force more than compensates for the 4.6% lower overall coefficient. There is approximately 0.5% more area in the 3-4 unit. It is apparent that if the triangular-pitch unit had the same air throughput as the 3-4 unit, the overall coefficient would have been considerably more than 4.6% better than the 3-4 unit coefficient. Also) the exit air temperature would have been lowered and the At driving force would have been increased correspondingly. This would have resulted in a considerable increase in the heat transferred by the triangular-pitch unit. Therefore, at the same -face velocity, the triangular-pitch unit would have outperformed the 3-4 unit, In effect, the triangular-pitch unit was seriously penalized by the lower air-face velocity since this resulted in a coefficient only 4a,6 greater than the 3-4 unit. In addition to this, the reduced air velocity resulted in a penalizing At driving force which dragged down the effectiveness of the coefficient still further. - 50

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN V. HORSEPOWER AND At CONSIDERATIONS A comparison can be made between the staggered-pitch unit and the three-in-line, fourth-row-staggered unit on the basis of ho vs theoretical horsepower, using the correlation report and the ALCO data. Figure 20 graphically presents the curves obtained. Sample calculations for preparing the curves are given in Appendix H. For reference purposes face velocity scales are superimposed on the figure. The upper face velocity scale is to be used only with the lines labeled "triangular." The lower face velocity scale is to be used only with the lower curve labeled "three-in-line, fourthrow-staggered, ALCO data." It should also be noted that Fig. 20 is limited to 2-in.-OD finned tubes having 7 fins per inch, since the ALCO data are for such a tube. Figure 20 can be used for making two comparisons, one based on equal face velocities and the other based on equal horsepower. Comparison 1 (same face velocity): assume Vface = 800 ft/min, (ho)^ = 16.4, (hp)A = 0.135 (ho) 4 = (ho)A 16.4 (ho)3,4 11.0 11.0, (hp)3,4 = 0.075 = 1.493 or 49.3% greater ho (hp)A (hp)3,4 0.135 0.075 1.80 or 80% greater. Comparison 2 (same horsepower): assume hp = 0.1 51

I Inn.I II,,-, -- - - - - - ---- 1I I II-IIT. I I --- -- - - - - - - -Face velocity (triangular pitch) _____ 0 60 11 e _800 K)1000 1200o 1400 1600 __ L J-..4,-f Triangular (h') _____ — _ —_ __ _ _ — Triangular (ho) (hn) 3 in. line - - o I z un o N) r I!0 ss;" t,a =t;u.~ _ 4th staggered (Alco data) I I I I. fI Wolverine tubes 7 fin/in. Do = 2" Dp 1.08" Ad= 2.83ft2ft S__S2 1/16 Fac velocity (3 in line 4th stoggered) 400 600 800 I 1000 1200 1400 1600 2000 I,,, J,I I I i [ ]1,. Mr1. m Z m z m m w I m m 0 > -n I 0 r' Cm -H m z 0 s" z.01.Q ID THEORETICAL HP/FT2 OF FACE AREA Fig. 20. Outside coefficients vs theoretical horsepower.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN (ho)A = 15.4, (Vface) = 720 (ho)3,4 = 11.6, (Vface)3,4 = 880 (ho)A (h0)3,4 15.4 = 1.327 or 32.7% greater for the same hp. 11.6 The ratio of volume of air moved by blower: (V)A (V)3,4 720 = o818 880 o and (1.000 - 0.818) =.182 or 18.2%, or, the triangular unit gives a 32.7% higher air-side coefficient but 18.2% less air is pumped with the same theoretical horsepower. The above discussion concerned only the controlling coefficient, ho, and the corresponding theoretical horsepower. The corresponding heat duties, Q, cannot be obtained directly from Fig. 20. The quantity of energy transferred is given by Q = UoAo (At)LM and Uo is a function of ho, hi, ro rij and rw. For discussion purposes, hi, ro, ri, and rw may be assumed to be constant. Taking hi = 1000 rO = 0 ri = 0.0125 based on outside area rw = 0.0005 based on outside area. Referring to comparison 2, above (same hp): (U)A - (l\4 + 0,0125 + 0.0005 + (10015-) = 0.09. * (Uo)A = 11.1 53

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN () 34 (= ( 6) ++ 0.0125 + 0.0005 + (128) 0.1112 (U0)3,4 = 8.95 (UO) 11.1 = 1.24 or 24% greater Uo (as compared (U0)3,4 8.95 with 32.7% on ho basis). Therefore, QA = Uo AA (At)A Q3,4 U3,4 A3,4 (At)3,4 for one sq ft of face area, AA = A3,4 QA =:1.24 (At Q3s4 (At)3,4 The fact that the triangular-pitch unit is handling 18.2% less air with a 24% greater overall coefficient will result in heating the air passing through the triangular-pitch unit to a higher exit air temperature than the air leaving the three-in-line, fourth-row-staggered unit. This will result in a lower log mean temperature difference driving force for the triangularpitch unit. This in turn will reduce the ratio of QA to Q3 4 to less than 1.24o The triangular pitch arrangement gives a higher ho and Uo than the three-in-line, fourth-row-staggered unit at the same theoretical horsepower input. 54

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN VI. PRESSURE-DROP COMPARISONS The predicted air-side pressure drop in inches of water for the 3-4 unit (Monday's data) is 0.248, using the ALCO curve (see Table V). The predicted air-side pressure drop for the triangular-pitch unit (Monday's data) is 0.356 using Fig. 16 (see Table VI). The value taken from the ALCO data as given in Table V is used because it is based on experimental test data. The values are given for the test conditions and therefore are not based on any ideal comparison basis. The Happy Co. measured 0.78 in. of water for the triangular-pitch unit and 0,54 in. of water for the 3-in-line, fourth-row-staggered unit (Monday's data, see Section X). Tables VIII and IX (see Section X) present the data obtained by the Happy Co. on Thursday, April 28, and May 2, 1955, respectively. The Happy Co. measurements included a static-pressure reading which was made between the fan and the tube bank. These measurements were made by inserting a plain copper tube underneath the tube bank, with one end open and the other end connected by means of rubber tubing to an inclined manometer. The values obtained by the Happy Co. are given in Table X where they are compared with the values computed from the available correlations. The values tabulated for Mondayts test are the test values for 2:25 p.m., as this time corresponds closest to the time during which the velocity profile was being made. TABLE X SUMMARY OF PRESSURE-DROP ANALYSIS Three in line, fourth row staggered (see Table V) AP, Inches of Water Thursday s test data (Vface = 303) (Happy Co.) East bay (avg) 0.325 West bay (avg) 0,245 average of bays 0.285 Predicted by the Happy Co. rating sheet 0.o96 55

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE X (Concl.) AP, Inches of Water Predicted by 651C 0.09 Predicted by ALCO data line (7 fins per inch) 0.115 Mondayts test (Vface = 475) Field test data (Happy Co.) (avg) 0.54 Predicted by the Happy Co. rating sheet 0.22 Predicted by 651C 0.202 Predicted by ALCO data line (7 fins per inch) 0.248 The design-face velocity was 600 ft/min. The predicted pressure drops for this velocity are given in Table XI. TABLE XI COMPARISON OF PRESSURE DROPS WITH PREDICTED VALUES FOR 600-FT/MIN FACE VELOCITY AP, Inches of Water Field test data (Thursday) 0.285 Field test data (Monday) 0.540 Predicted by Happy Co. rating curve 0.344 Predicted by 651C 0.315 Predicted by ALCO data (7 fins per inch) 0.370 Tables X and XI indicate that the Happy Co. pressure-drop measurements do not agree with any of the predicted values for design or test flowrate conditions. 56

7- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - VII. FAN PERFORMANCE The fans used in the field-test units were manufactured by the Moore Co. of Kansas City, Mo. The fans are classified as series 48, 16-ft, 4-blade Moore Pressure Blowers, 7248-AM-16-4, 250 rpm. The Happy Co. furnished "anticipated-performance" curves for the fans, a copy of which is given in Fig. 21. The maximum fan pitch indicated on the figure is 15.6~0 Figure 21 gives the anticipated static pressure in inches of water and corresponding brake horsepower input to the fan as a function of volume of air handled with fan pitch as a parameter. The average face velocity measured on Thursday was 303 ft/min (p = 0.074) which corresponds to 345 ft/min (p = 0.065). The face area of 366 sq ft, when put together with the face velocity of 345, gives 126.2 thousands of CFM (see Fig. 21). The predicted static-pressure drop was 0.096 in. by Happy Co. and 0.09 in. by 651C for an average of 0.093 in. of water. These values, when plotted on Fig. 21, do not fall near the curves. The average face velocity for the three-in-line, fourth-rowstaggered unit (Monday's test data) was 475 ft/min (p = 0.074), which corresponds to 541 ft/min (p = 0.065). This corresponds to 198 thousands of CFM. The predicted static-pressure was 0.22 in. of water by Happy Co. and 0.20 in. by 651C for an average of 0.21 in. These values, when plotted on Fig. 21, do not fall near the curves. The average face velocity for the triangular-pitch unit (Monday's data) was 398 ft/min (p = 0.074), which corresponds to 453 ft/min (p = 0.065). This in turn corresponds to 166 thousands of CFM. The predicted static pressure by the correlation report was 0.356 in. of water. These values, when plotted on Fig. 21, do not fall on the curves. Blade-efficiency curves can be prepared, using the anticipatedperformance curves of Fig. 21. Sample computations are presented in Appendix L and resulting curves are presented in Fig. 22. Superimposed on Fig. 22 are vertical lines corresponding to the predicted pressure drops for the units, based on Monday's test data. Also indicated on Fig. 22 are two points corresponding to the Happy Co. field-test static-pressure measurements. The figure indicates that the predicted static pressures fall within the curves in the range of 52 to 45% blade efficiency. The peak of the curve reaches a maximum of 55% efficiency. It is believed that the existing fans are incapable of ever providing 600 ft/min face velocity air to the units tested. 57

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Anticisated perfortnnce Series 48, 16 foot, 4 Blade Moore Preas.ere Blower, 724&-AN-6-64 25C R.,.065 rounds per cubic foot air density 4s "40 20 10 o.5 C) r 4.1 0................. ~..........' i - I - - W - t.............. _..;......F_ __.... + -._ -- I............ -. -.1. " 2'... - 10 --.._..._ _. I'...,...*.-%_. \. 4....!b.....,,,.... Blade angle, in degrees, at the..... -/ 2 ioot dianeter point..,.. I Ni. \ - i,*.. *. _..>.... _ \'.{,.. ——.-........... \....1 _ k _ L 17 8 _ L. 6 i 0 100 200 300 Thousands of CFM 400 500 fA w40 16-Z*.;0. DATE REVISION —-.. -,...-;O. cCAT _..~_ REISION_~ For Atlantic Refining Compaqry Lovington, New Mexico............ SCALE THE.-.-. - - - ------- --------- -DRAWtC BYT.: CO, |_ ______,.__|___JO____|_________ o__KANSAS Ct, O. Jo, " 66-263 we. NS 1SL3 0, Fig. 21. Jacket-water-cooler anticipated fan performance curves. 58

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN r I iJ rn m lZ =1 STATIC PRESSURE, INCHES OF WATER Fig. 22. Predicted blade efficency of jacket-water-cooler fans. 59

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 68 64 AlI aluminum tube 60 56 52 48 Tube 2after cleaning 44 uJ i^ i ^' Tube I after cleaning w 40 O 36 Cl) / {, Tube I before cleaning I - 28 co /__ - i I WATER VELOCITY INSIDE TUBES, FT/SEC Fig. 23. Results of tests conducted on tubes cut from the jacket water cooler.

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Difference of resistances = 1 - 1 = 0.0175 25.0 44.4 ri = 0.0175, based on outside area, or ri = 00175 = 0.00128, based on liner area. 13-7 (Note: Any bond resistance present would cancel out.) Also: ri 0,017i = 0.00115, based on inside area. 15.2 Also given in Fig. 25 is a test curve for an all-aluminum tube which was tested at the same time. This curve is useful for predicting the amount of bond resistance presumably existing in the tubes. To indicate how this curve can be used, Fig. 24 was prepared. Figure 24 is based on the allaluminum test curve of Fig. 23. and gives the decrease in heat transfer as indicated by the percentages when the outside coefficient is 10. Figure 24 has a superimposed dashed line which represents the average performance of tubes No. 1 and No. 2 after cleaning, as given in Fig. 23. It is apparent that the bond resistance present amounts to approximately 5% of the overall resistance. This degree of resistance cannot account for more than 5% of the field-performance discrepancy as compared to the specifications for the units tested. One of the tubes tested was shipped to Wolverine tube for expanding of the liner. After the liner had been expanded, the tube was retested under identical test conditions. The results of the expanding are shown in Fig. 24 by a dash-dot line indicating reduced bond resistance from 5% to 4%. The tubes used in the fabrication of the units that were tested on Monday had expanded liners. 62

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - F-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I W 40 LIUJ 36 36 ------ y — 7-7 / --- ---------—.^ —- 10% Reduct "L 32 IL Do 2 20'. 20 /30~ 1V/ 2 — -- -- --- -- - 40' Assumed hQ= 10.0 based on outside /ffa - -area 0 2 4 6 8 10 12 14 16 V+, FEET/SEC Fig. 24. Predicted performance of tubes tested in field application. 63 I

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - IX. JACKET WATER TEST-ANALYSIS CONCLUSIONS The conclusions drawn from the analysis given in Sections II through VIII of this report are: 1. The units are operating at air-face velocities that are well below the design-face velocity of 600 ft/min. 2. The design air-side coefficients for three-in-line, fourth-rowstaggered units are greatly overrated. The air-side coefficient controls the performance of the unit. 3. A combination of the above two factors explains the poor performance obtained from the units. 64

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN X. HAPPY CO. TEST DATA The representatives of the Happy Co. took test data of their own on Thursday and on Monday, April 28 and May 2, respectively. Their Thursday test data are given in Table VIII and their Monday test data are given in Table IX. TABLE VIII HAPPY CO. TEST DATA OF THURSDAY, APRIL 28, 1955 Air temperatures, ~F: West Fan (3:43 p.m.) 77.0 76.0 76.5 76.5 76.0 (3:43 p.m.) 75.0 77.5 75.0 7-5.0 74.0 East Fan (4:40 p.m.) 75.0 77.5 75.0 74.0 75.0 Water temperatures, ~F: Coil 1 Inlet Inlet Out let Outlet Outlet Outlet 164.2 164.4 157.3 157.6 Coil 2 164 164 156.2 157.1 157.4 156.6 Coil 5 164.2 164.0 155.9 155.0 Coil 4 164 156 Water-main inlet temperature Water-main outlet temperature = 164~F = 156~F 65

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - Static pressures: West Fan 0.26 0.23 East Fan 0.33 0.32 = 18.5 = 19.0 Motor amps: West Fan East Fan Pump suction: 4.5 psi Pump discharge: 25.5 psi Water-main Pitot tube: 3.5 to 8 in, of meriman oil, specific gravity = 2.9. (readings fluctuated widely) Air pressure to fans at 4:40 p.m. (controller): 9.0 psi. 66

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE IX HAPPY CO. TEST DATA OF MONDAY, MAY 2, 1955 Time 1:45 p.m. 2:25 p.mo Water Water Water Water Water Water In, To, ~F Out, T1, ~F Out, T2, ~F Out, T3, ~F Out, T4, ~F Out, T2, ~F 154.7 147.2 145.5 145.8 144.9 146.0 149.7 142.1 140.9 141. 140.3 141.3 (Overall) Flow Meter, in H20 Pump Suction, psig Pump Disc, psig Air Wet Bulb, ~F Air Dry Bulb, ~F Barometer Wind Direction Wind Velocity, Mph, est. 65 avg 2.8 22.0 66 avg 2.75 22.0 67 74 WSW 12 WSW 12 Fan Position, (Air Pressure) Static Pressure, West Unit Static Pressure, East Unit Air Temp., ~F, West, in Air Temp., ~F, East, in Air Temp., out, Avg, ~F, W. Air Temp., out, Avg, ~F, E. Air Velocity, Avg, FPM, W. Air Velocity, Avg, FPM, E. Fan Load, amps, West Fan Load, amps, East Fan Volts Speed, rpm. West Speed, rpm. East O(15,6~) 0.75-0. 85 o.48-o.56 81 81.5 0(15.60) 0.78 0.53-.55 79 79.5 40 40 470 at Gen. 250 250 42 42 470 at Gen. No. Compressor on. Discharge Press, psig Suction Press, psig Flowmeter, Diff, "H20 Temp. Gas In, ~F No. 1 Generator Load, KW No. 2 Generator Load, KW 5 880 6 35 69 390 390 5 885 6 35 69 350 320 67

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN -l XI. HAPPY CO. SPECIFICATION SHEETS The specification sheets for the five Happy Co. units at the Atlantic gasoline plant are reproduced for reference purposes on the following pages. 68

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN COMPANY TULSA, OKLAHOMA SPECIFICATIONS FOR L HAPPY FIN-TYPE COOLERS Customer DreaSer gnin'ering C mpny Quotation No. Address P 28 Tulsa; Oklom Date a 6 953 Plant Atla ic Refining, Lovingtono N M_.Reference Tt_ 2.1 A..l _r_ Cooler Model HFR-360-2 Service Jacket Water....__.___ INSIDE TUBES' I AIR DATA Fluid __ Elevation, Feet 3700 Sp. Gr. @ 60F. Inlet Air Temp., ~F. 100 Sp. Ht. @ Avq. Temp, Total SCFM 445,000 Density @ Avg. Temp. SCFM/Fan 500 iscosity Temp. Rise, F. 44.8 xiMcM M.T.D, Based on bare tube 30' MECHANICAL EQUIPMENT U based on bare tube 159.5 Fans: Type loore 048 Adij Pit h ____I__ _Inside fouling factor 0.001 No. 2 Qty.-Gas Dia. 161 H. C. Vapor, #/Hr. No. Blades 4 Steam, #/Hr. RPM 250 H. C. Liquid, t/Hr. -. Hp./Fan 29.9 Water GP 3R6 Total Hp. 59.8 Gears: Make C leveland nlet Temp, ~F. 161,06 Model CU40 utlet Temp., ~F. 150 Rating AGMA 47 perating Press., PSI 12 Ratio 7:1 Pressure Drop, PSI 4,4 5 Jt Drive: _ Design Pressure, PSIG 5i Supported rest Pressure, PSIG1 o _.. sla n _ Design Temperature, OF. 1 75 Heat Exchanged, BTU/Hr.21 49200 Surface, Sq. Ft. Totalbare 45 total external 62A000 Motors: Make TLois Allisa Surface per Coil, Sq. Ft. Total bare 1125; total external 15,502 Type Explosion proof Number __ MATERIALS & CONSTRUCTION Hp. 40 Units Reqd. __R.P.M. 1760 Coils per Unit 4Volts 440 Coil Width 7.5t Phase 3 Tubes per Coil 179 Cycle 6 n Tube Length 24t Gas Eng.: Make Tubes i' olverine[ 1" x 18 BWG Inhibited Admiralty Liner Model Fins Extruded Aluminum 9/inch, 1/2 high.019" _ Number Headers Box Fabricated Steel, Brass Plug Hp. Connections 8" - 150# R F. R.P.M. No. Passes 2' k I J I Cooler made up of 2 bays, 1 fan per bay, each bay approximately 16' x 24' x 121. Both hays combined in common st ructure. 69

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - 7/u l^ J COM PANY TULSA, OKLAHOMA SPECIFICATIONS FOR L HAPPY FIN-TYPE COOLERS Page 2- Alternate Customer Dremar Engineeri Company Quotation No. 9117 Address Tulsa OkahomaDate March 11 1953 Plant Atlatic Refinlag Co.. Lovington, N.M. Reference Item 2-2 Alternate Cooler Model HFR-72 Service Evaporator Condenser, ______IDINSIDE TUBES AIR DATA Fluid_ Elevation, Feet 3700 Sp. Gr. @ 60'F. Inlet Air Temp.,'F. 100 Sp. Ht. @ Avg. Temp. Total SCFM 46.200 Density ( Avg. Temp. _SCFM/Fan 23100 Viscosity Temp. Rise, ~F. 79.2 stckOx Ml.T.D, ased on bare tube surface 78 MECHANICAL EQUIPMENT -I__ T.od oan ha.re taZbe 17;2 __ FFans: TypeKoppers Adj. Pitch InAide_ fouling f acir. ntn I,No. 2 Qty.-Gas Dia. 5t H. C. Vapor, #/Hr. No. Blades 4 Steam, i/Hr. Suated 4.00... - RPM 870 H. C. Liquid, #/Hr. _ Hp./Fan 3,3 Water Total Hp. 6e6 Gears: Make nlet Temp, ~F. 240 Model utlet Temp., ~F. 200 Rating AGMA perating Press., PSI' 25 Ratio Pressure Drop, PSI 2 [II__l Fan direct Design Pressure, PSIG 50 C onnected to rest Pressure, PSIG 100 S aX motor esign Temperhture, ~F. 30V-Belts Heat Exchanged, BTU/Hr. 3 970.000 Surface, Sq. Ft. Bare 298.5 Total External 4120 Motors: Make Louti Allis urface per Coil, Sq. Ft. 2 98.5 4120 Type TEFC" ___________________________________ Number 2 MATERIALS & CONSTRUCTION Hp. 5 Units Reqd. R.P.M. 8 Coils per Unit 1 Volts 440 Coil Width 61 Phase 3 Tubes per Coil 95 - 3 rows Cycle 60 Tube Length] __Gas Eng.: Make Tubes Model Tubes Wohlvertne,IIIx lat WG Inhibited Admiralty Liner,Model Fins Extrudd A Lminum 9/inch, 1/2" high.019' Number Headers Lox Fabiicated Steel Brass Plug Hp. Connections 6" - 150# RF. In: 2" -150# R.F. Out R.P.M. No. Passes 2 - reducing 66-2/3 - 33 -1/3% 70

..- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 7A~li~..^..... OMPANY TULSA, OKLAHOMA SPECIFICATIONS FOR HAPPY FIN-TYPE COOLERS Customer De r ginering Compay Quotation No. 9 14 Address Tulsa, Oklahorma___ ______Date March 10 1953-Q Plant Atlantic Refining Co.. Lovington. N.M. Reference i Altern Cooler Model HR40 Service Still Overhead Partial Condenser _________________________ I.........INSIDE TUBES...AIR DA.TA... Fluid_ Gasoline Vapor Elevation. Feet 37 Sp. Gr. @ 60'F. _ Inlet Air Temp., ~F. 10 - Sp. Ht. {_ Avg. Temp. Total SCFM 258,00 Density (a, Avg. Temp. SCFM/Fan 129,000 Viscosity _Temp. Rise, ~F. 2'7 7 M*k3c. MT.TD bare tube 46.0 MECHANICAL EQUIPMENT _ T based on I-are L_ ___68.2ube____ Fans:' Type Ore Moore Inside fouling fator_____ INo. 2 Qty.-Gas_ Dia. 11 H. C. Vapor, #/Hr. n 74,4,M.W. 58 4 No. Blades' Steam, #/Hr. M.W. 58.4 RPM_ 8 ------'- /J —---- 31 _.._ _ —-— ~-r- -............... —..- _._- _ PM 18_.. H. C. Liquid, #/Hr. ____ - Hp./Fan _ WaterOt 2,072 T. Total Hp. 39 8 ______Ass str^^gh i~ ^Gears: Make Clec- la d nlet Temp,'F. 180 Model utlet Temp., ~F. 1 Rating AGMA 2-.... perating Press., PSI ____Ratio 5-4A / 1 Pressure Drop, PSI2 Ger & 2otor - Design Pressure, PSIG Q___e n C Onrete Test Pressure, PSIG 225,_ec. P rdea Design Temperature,'F. 00| Heat Exchanged, BTU/Hr. 7,747 000 Surface, Sq. Ft. Bae 472 t 34050 Motors: Make Louis-Alli_ Surface per Coil, Sq. Ft. 11 Type yp. Prof Number 2 MATERIALS & CONSTRUCTION Hp. //, Units Reqd. -R.P.M. 1800/90 Coils per Unit, Prll Volts 440 Coil Width A1 Phase Tubes per Coil 131 - 4 rows Cycle 60. - Tube Length 24t Gas Eng.: Make Tubes Wov]~n Model ~~~Tubes WO tfi[t1, x 18 RWG!nhihited Admiralty Tiner Model Fins Extruded Al minum 9/inch 1/2" _.019 Number Headers Fabric d Steel, Beass Plu Hp. Connections 8"15 R.P.M. No. Passes -...... Bay Size Approx. 20t x 241 x 11' 1 71

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - OMPANY TULSA, OKLAHOMA SPECIFICATIONS FOR HAPPY FIN-TYPE COOLERS Customer Dresse. Engineering Company Quotation No. 9 Address Tulsa OklahomaDate Plant Atnic Refining Co. VNGTON. N.M. Reference Item 2-4 Alternmte Cooler Model HFR-180 Service Debutanizer Overhead Condenser! ________INSDE TUBESTU AIR DATA. Fluid _ ____ VaporElevation, Feet 3700 - Sp. Gr. @ 60'F. Inlet Air Temp., ~F. 1 Sp. Ht. @ Avg. Temp. Total SCFM 12 1,010 ensity (, Avg. Temp. SCFM/Fan 62,500 iscosity Temp. Rise, ~F. 35.6 Mol. Wt. 58 MECHANICAL EQUIPMENT Fouling Factoo 0.001 Fans: Type U based on bare tube surface 104.9 No. Qty.-Gas _Dia. 71 H. C. Vapor, #/Hr. IN: 33,859 No. Blades I _ St eam, #/Hr. RPM __557 ____ H. C. Liquid, #/Hr. OUT: 33t859 Hp./Fan 12 Water Total Hp. 25.6..___._._____ —___.....____ ~Gears: Make C leveland Worm nlet Temp, ~F. 174' Model C U 1000 utlet Temp.,'F. 140. Rating AGMA 1 j 7 rp Operating Press., PSI 158 _Ratio 3.117 to 1 Pressure Drop, PSI 2. W: ear and r O0 Design Pressure, PSIG 175 _ r'_= rest Pressure, PSIG 265 pedestal f~i~.,N pedestal Design Temperature,'F. 200 V-Belts Heat Exchanged, BTU/Hr. 4,800,000 Surface, Sq. Ft. Bare 1125 Total.External 15,500 Motors: Make Louis Allis Surface per Coil, Sq. Ft. 1125 Total External 15,500Expl proof2spe Number 2 ________ MATERIALS & CONSTRUCTION Hp. 15/3 75 Units Reqd. R.P.M. 1 7 /875 Coils per Unit 1 Volts 44 Coil Width 7 51 Phase 3 Tubes per Coil 179 - 4 rows Cycle 60 Tube Length 24:Gas Eng.: Make Tubes Wolverite 1, OD x 18 BWG Inhibited Adniralty Model Fins Extruded S ooth Aluminum. 9/inch, 1I2 x 0.019" Number tteades Hp. Headers _ __- Fabricated Steel, Box Brass Plug Hp. Connections In 8" 150 R.. Out 6" o 150# R F. R.P.M. No. Passes 2 Reducing 75 25 Z.8 Z R e, 757 -e 2u__ 5.....5 I MT''V = 40.U Daseo on straignt Line neat release, Composttton point so this should be,a very conservative selection. indtcates a 109" bUDbbl 72

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - TUA. O C[OMPANY TULSA, OKLAHOMA SPECIFICATIONS FOR L HAPPY FIN-TYPE COOLERS Pae o5 Customer e Quotation No. 9117 Address p p Q 518; TnIa k Date March 4, 1953 Plant AtA a Rofining Companyn. lvington. New virno Reference Item 2-5 Cooler Model HFR -60 Service Lean Oil Cooler i__ ______INSIDE TUBES AIR DATA Fluid Lean Oil ElevatiOn, Feet 3700 Sp. Gr. @ 60'F..825 Inlet Air Temp., ~F. 100 Sp. Ht. @ Avg. Temp. Total SCFM 46,500 ensity @ Avg. Temp. SCFM/Fan 46,500 iscosity.g F 1.49 Temp. Rise, ~F. 30,5 MoL Wt 200 MECHANICAL EQUIPMENT Fouling factor 0.002 Fans: Type Moore 24, Axial U based on bare tube surface 98.A No. 1 Qty.-Gas -_Dia. 7 H. C. Vapor, #/Hr. No. Blades 5 Steam, #/Hr. RPM 557 H. C. Liquid, #/Hr. 0.6_ _ Hp./Fan 10.5 Water Total Hp. 10.5 MTD 41.2 Gears: Make Cleveland Worm nlet Temp, "F. 173 Model C U. 100 utlet Temp., "F. 140 Rating AGMA 15.7 HP perating Press., PSI 46 Ratio 3-1/7 to I Pressure Drop, PSI 10 V-Belt Drive: esign Pressure, PSIG 125 Fan Sheave Test Pressure, PSIG 190 Drive Sheave esign Temperature, F. 250 V-Belts eat Exchanged, BTU/Hr. 1,538 000_ Surface, Sq. Ft. Bare tube 376, Total external 5180 Motors: Make outs Allis Surface per Coil, Sq. Ft. Bare tube 376. Total external 5180 Type Expl. proof 1-s8.____"__________________________ Number 1 MATERIALS & CONSTRUCTION Hp. 15 Units Reqd. 1 R.P.M. 1750 Coils per Unit 1 Volts 440 Coil Width 7.5! Phase 3 Tubes per Coil 179 - 4 rows Cycle 60 Tube Length 8 Gas Eng.: Make - Tubes Wolverine l"x 18 BWG Inhibited Admiralty Model Fins Extruded, S ooth, 9/inch, 1/2" x.019"Aluminum Number Headers Fabricated Steel, Brass Plug Hp. Connections 61" 150# R.F. R.P.M. No. Passes 10 low >ee< I 1 73

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN XII. ANALYSIS OF UNITS NOT TESTED A. LEAN OIL COOLER The specifications for this unit are given in Section XI. This unit had been field tested during 1954 by the Happy Co. and the Dresser Engineering Co. The results of their tests indicated that this unit was performing at about the same level as the jacket water cooler. The jacketwater-cooler performance, as indicated in Sections II and III of this report, was due to the air-side coefficient being considerably lower than that given by the Happy Co. rating sheet. Therefore, the lean-oil-cooler design was evaluated in terms of the ALCO data converted to 9 fins per inch. This analysis is given in Appendix I and is summarized below. The design air flow rate of 46,500 std cu ft/min as given in the specifications was assumed to be that delivered by the fans. The corresponding face velocity is computed as 760 std ft/min. The design heat duty as given in the specification sheets for the unit is 1,538,000 Btu/hr. TABLE XII PREDICTED PERFORMANCE OF LEAN OIL COOLER (Based on ALCO data converted to 9 fins per inch, assuming fan produces 760 ft/min face velocity) Vface, std ft/min Q, Btu/hr Q/Q1esign Uo (based on outside area) (Uo)L (based on liner area) Uo/Uo (specified) toutlet oil, ~F (ho)L (liner area) ho (outside area) hi (based on inside area) rfouling inside rmetal ATM, ~F 760 1,210 000 79% 5.49 66.9 71% 147.5 133 9.7 216 0.002 0.00062 48.1 74

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN For comparison purposes, similar calculations were made, using the outside air film coefficient obtained from 651C. The analysis is given in Appendix I and is summarized in Table XIII. TABLE XIII LEAN-OIL-COOLER ANALYSIS BASED ON 651C AND Vface OF 760 STD FT/MIN Vface, std ft/min Q, Btu/hr Q/Qdesign Uo (based on outside area) (Uo)L (based on liner area) UO/Uo (specified) toutlet oil, ~F (ho)L (liner area) ho (outside area) hi (based on inside area) rfouling inside rmetal ATM, ~F 760 1, 450,000 935 6.4 87.8 89% 142.5 252 18.4 216 0.002 0.00062 43.2 For comparison purposes, similar calculations were made, using the outside air film coefficient obtained from the Happy Co. rating curve. The analysis is given in Appendix I and is summarized in Table XIV TABLE XIV LEAN-OIL-COOLER ANALYSIS BASED ON HAPPY CO. RATING CURVE AND Vface OF 760 STD FT/MIN Vface, std ft/min Q^ Btu/hr Q/Qdesign Uo (based on outside area) Uo (based on liner area) Uo/Uo (specified) toutlet oil F ho(liner) hi (inside area) ri rmetal ATM, 0F 760 1,390,000 90.5% 6.08 83.5 85% 145.5 220 216.002.00062 44.2 75

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The specified heat duty of 1,538,000 Btu/hr cannot be realized by use of the unit specified. If the face velocity of 760 ft/min is obtained from the fan specified, the unit could be expected to perform at about 71% of design demand. On the basis of the Happy Co. rating curve, the unit would be 9.5% low on heat duty and 15% low on overall coefficient if the rating curve was correct. B. DEBUTANIZER OVERHEAD CONDENSER The specification sheet for this unit given in Section XI calls for a heat duty of 4,800,000 Btu/hr with an inlet hydrocarbon temperature of 174~F'and an exit hydrocarbon temperature of 140~F. The design inlet air temperature is specified as 100~F, with an air temperature rise of 35.6~F. The computed bubble point of the hydrocarbon mixture as given in Appendix J is 172~F and the dew point is 174~F. Therefore, the inlet mixture is saturated vapor which condenses over a 2~F range. The condensate must subcool 32~F, from 1720 to 140-~F. The computed condensing load on this basis is 4,070,000 Btu.hr and the subcooling load is 724,000 Btu/hr for a total of 4,794,000 Btu/hr. This checks the specified heat load. A major consideration in this design is one of insuring that the subcooling load will be obtained. The proposed design does not provide any means for controlling the condensate subcooling film coefficient. It should be emphasized that a partially flooded tube will have (a) a low condensate subcooling film coefficient resulting in high film resistance to heat transfer and (b) a tube-wall temperature that will approach the condensing-vapor temperature. In order to insure that the subcooling will be realized the bottom row of tubes should be completely flooded. The propos:ed design had two passes on the tube side with the top three rows as the first pass and the bottom row as the second pass. The condensing coefficient was computed to be 300, based on the inside surface area. The air-side coefficient.as determined by the ALCO data converted to 9 fins per inch was 9.28 for a face velocity of 700 std ft/min. This results in an overall condensing coefficient of 5.735 based on the outside area. The mean temperature difference for the condensing zone is 49.7~F. Using the condensing heat duty of 4,070,000 Btu/hr and the above data, a condensing surface area of 13,100 sq ft (outside area) is required. The top three rows of the proposed design provide 11,5300 sq ft or a deficiency of 1800 sq ft, which amounts to 1357%. Considering the subcooling zone to be the bottom row of tubes completely flooded, a condensate cooling film coefficient of 86.4, based on the inside area, is computed. If the air-side coefficient is taken as 9.28, the same as above, and is combined with the inside coefficient of 86.4, an 76

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN overall coefficient of 3.11 is obtained, based on the outside area. The temperature-difference driving force for this zone is 52.5~F. The required subcooling heat transfer area, based on the subcooling load of 724,000 Btu/hr, is 4380 sq ft of outside area. The bottom row actually provides only 3800 sq ft. The deficiency of 580 sq ft amounts to 1353%. Since no provisions were made for flooding the bottom row of tubes, the bottom row will be only partially flooded. This will result in the exposing of additional condensing surface area to the condensing vapors. Therefore, the 13.7% deficiency of condensing area will be eliminated, resulting in the condensation of the required condensing load. The required degree of subcooling will not be obtained under this arrangement. The analysis given above, based on ALCO data, indicates that the proposed design will not provide the required area for effecting the specified heat transfer. A similar set of calculations based on 651C can be made for comparison with the above analysis. Only the air-side film coefficient and At will be greatly affected. The air-side film coefficient was determined to be 16.6, based on outside area. The overall coefficient for the condensing zone was computed to be 7.86 with a At of 49.7~Fo This results in a required condensing area of 10,520 sq ft (outside area). The top three rows provide 11,300 sq ft, or an excess of 7o6%. The overall coefficient for the subcooling zone, assuming the bottom row is flooded, was computed to be 5*95, based on the outside area. The At was computed to be 52.2~F.. The corresponding required external surface area is 3800 sq ft. The bottom row provides 3800 sq ft. The analysis given above, based on 651C, indicates that if the bottom row were flooded, the unit would handle the specified heat duty. This assumes that the face velocity of 700 SFM is provided by the fan. The air-side coefficient as given by the Happy Co. rating curve at 700 std ft/min is 220, based on the liner area. This corresponds to 16.05, based on the outside area. This compares to 16.6, based on 651C. The condensing-zone overall coefficient becomes 7.73, based on outside area. This is l16% lower than the 651C value of 7.86. Therefore, the required condensing area is 10,720 sq ft. This compares with 11,300 sq ft provided by the top three rows for an excess area of 5.4%. The overall coefficient for the subcooling zone, based on the Happy rating line, is 3591, based on outside area. The required subcooling area becomes 3840 sq ft as compared to the 3800 sq ft provided by the bottom row. I D - - 77

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - r I I The field tests conducted on the jacket water cooler essentially substantiate the converted ALCO data for predicting the air-side coefficient for three-rows-in-line, fourth-row-staggered'arrangement. The most reliable prediction of performance of the debutanizer condenser is that predicted by the converted ALCO data. The Happy Co. rating curve and 651C both overrate the air-side coefficient. It is therefore concluded that the proposed design is not capable of providing both the required condensing load and the required subcooling load. C, EVAPORATOR CONDENSER The specification sheet for this unit given in Section XI calls for a heat duty of 3,970,000 Btu/hr with an inlet steam temperature of 2400F and an outlet condensate temperature of 200~F. The design inlet air temperature is specified as 100~F with an air temperature rise of 79.2~F. The required heat duty consists of two parts: (a) that of condensin 4000 lb of saturated steam per hour and (b) that of subcooling the resulting condensate 40~F. These heat loads are 3,808,000 and 160,000 Btu/hr, respectively. The proposed design provides a total external area of 4120 sq ft for accomplishing both the required condensing and the subcooling load. The specification sheet calls for an MTD of 78~F. If the inlet air temperature is 100~F, the exit air temperature is 179.2~F, the steam inlet temperature is 2400F, and the condensate outlet temperature is 200QF, a LMTD based on these terminal temperatures of 78.7~F is obtained. There is an implication that the use of this conservative MTD, when employed in conjunction with some overall coefficient, will give the proper heat transfer area. The problem here is one of determining the applicable overall coefficient. The overall coefficient given in the specification sheet is 172, based on liner area, or 12.55, based on the outside area. It can be readily shown that the overall coefficient of 12.55 is considerably larger than the applicable overall coefficient. The heat exchanger consists esentially of two zones, a condensing zone and a subcooling zone. Each zone has its effective overall coefficient and corresponding LMTD driving force. An overall coefficient for the condensing zone can be computed from the data given in 651C, Table 12, page 9. This overall coefficient per foot of tube is 50.5; this assumes no fouling. This value corresponds to 14.05, based on outside area. Allowing for inside fouling of 0.001 reduces this value to 11.6 for the condensing zone. This value is 8.6% lower than the Happy Co. value of 12.55 for the entire unit. I a ] - - 78

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The subcooling-zone overall coefficient will be considerably lower than 11.6. It must be emphasized that it is necessary to insure complete flooding of some tubes in order to obtain the degree of subcooling specified for this unit. If the tubes in the subcooling section are partially filled with condensing steam, the tube-wall temperatures will approach those computed in Appendix K. It is shown in this appendix that if steam is present in the tube, it is impossible to attain the specified outlet steam condensate temperature of 200~F. If the bottom row of the evaporator condenser is completely flooded, the overall subcooling coefficient for this zone can be shown to be 4,535 based upon the air film coefficient given by 651C. The overall subcooling-section coefficient of 4.53 can be combined with the condensing-section overall coefficient of 11.6 to give an effective overall coefficient of 11.0 for the entire unit. This mean overall coefficient is obtained by combining the overall coefficient for each section with its respective At and area. The mean overall coefficient of 11.0 predicted by 651C is 12.3% lower than the Happy Co. value of 12.55. Since the jacket-water-cooler field test results indicate that the air film coefficient predicted by 651C is too high, it is believed that the unit is not capable of producing both the required condensing load and the required subcooling loads The ALCO data converted to 9 fins per inch are limited to the threerows-in-line, fourth-row-staggered arrangement. These data do not apply to a two-rows-in-line, third-row-staggered arrangement. Therefore, no analysis was made, using these data. Do STILL OVERHEAD PARTIAL CONDENSER The specification sheet for this unit calls for a heat duty of 7,747,000 Btu/hr with an inlet hydrocarbon and steam temperature of 180~F. The exit vapor-liquid mixture is to be 140~F. The molecular weight of hydrocarbons entering is 58.4 and the molecular weight of condensing material is 693., The air to the unit is 100~F with a temperature rise of 27.7~F. In order to evaluate the proposed design for this application, the feed-stream composition used as the basis for the design must be available. The only composition available was that provided by the Happy Co. as the result of a field test made on August 11, 1954. Whether or not this composition checks with the design composition is not known. A series of calculations based on this available stream composition was made. First, a check of the molecular weight of the feed stream was made and was found to be 57.45, as compared to the 5864 value given above. The computed molecular weight of the condensed material was calculated as 6800, as compared to the 69.3 value given above. These computed values appear to check fairly well 79

r- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - with the specification-sheet valueso The computed heat duty, using the specification sheet flow rates and the field test analysis, is 5,729,000 Btu/hr, as compared to 7,747,000 Btu/hr given in the specificationso Calculations indicate that all the steam does not condense in this unit. This accounts for part of the reduced computed heat dutyo The feed stream contains eight components, C2> C3, iC4, nC4, iC5, C,. C7, and steam. In flowing through the heat exchanger, about 29% of the material condenses (based on field analysis). This compares with the specification sheetts computed condensing value of 37.5%. The mechanisms involved in partial condensing are complex. The condensing material must diffuse to the condensing surface by means of molecular diffusion or eddy diffusion. The remaining gas stream must cool by convective heat transfer. The complex mechanisms involved when eight components are partially condensed simultaneously at varying rates throughout the exchanger are such that exact rate analysis can be made only by use of electronic digital computers. Therefore, recourse must be made to the use of empirical short-cut methods. In the case at hand, the condensing film coefficients will be very large in comparison with the hydrocarbon-gas film cooling coefficient. The air-side coefficient will be very low and the combination of hydrocarbon-gas phase cooling resistance with the air-side film resistance will control the performance of the unit. An evaluation of the unit can be made on this basis by proper distributions of coefficients. The computed gas film cooling resistance is frequently reduced by multiplying by the ratio of the gas cooling sensible heat duty to the total heat transfer heat duty. An analysis of the proposed unit was made on this basis. The unit designed contained two passes on the tube side. The inlet top pass contained three rows of tubes and the bottom pass contained one row of tubes. Inside gas film coefficients were computed for both the upper and the lower passes at the average temperature, The values of hi obtained were 62.5 and 92 for the upper and lower pass, respectively. As can be seen from the numerical values of these coefficients, the effect of the decreased flow area in the lower pass as compared to the upper pass is to increase the inside coefficient for this region, even though the vapor-mass flow rate has been reduced through condensation of part of the vapor. An analysis using the inside gas film coefficients given above, combined with an air-side coefficient computed from 651C, was made, This analysis resulted in a predicted required area for the design conditions and the field-test feed-stream composition of 23,100 sq ft of external area. The area provided in this unit is 34,050 sq ft for an excess area of 48% over that required. A similar analysis, using an air-side coefficient given by the converted ALCO data, predicts a required area of 33,200 sq ft (external) as 80

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - compared to the provided area of 34,050 sq ft. This amounts to 2.5% excess area over that provided. In both the above analyses the condensing film coefficient has been ignored. The diffusion-rate phenomenon also has been neglected. The assumption has been made that the gas film coefficient which is controlling can be increased due to the fact that the higher condensing coefficient results in a greater overall effective coefficient. In the case under consideration, the condensing load is about 75% of the total heat duty. This would result in a much larger effective overall coefficient than the hydrocarbon-gas film coefficient alone would indicate. On the basis of the above analysis, the proposed design would be expected to give the specified outlet hydrocarbon temperature when the specified mass of material of field-test composition is fed to the unit. Results of tests made by the Dresser Engineering Co. and the Happy Co. on August 8, 1954, indicated that this unit had a performance index of 65%. This discrepancy with the calculated values, using the converted ALCO data, could be due to one or more of the following reasons: 1. The field-test composition and pressure of the inlet vapor mixture for this test indicates a dew point of approximately 1820F as compared to the inlet feed temperature given as 192.8~Fo This indicates that an appreciable amount of superheat was present in the feed gas mixture. The removal of the superheat prior to the start of condensation was not provided for in the design and would tend to reduce the performance index of this unit. 2o It was determined that the heat duty specified was based on the assumption that all the steam entering the unit would condense. Calculations indicate that approximately one-fourth of the steam entering the unit would not be condensed. This would cause the heat load of the unit to be less than the value given in the specifications when the composition flow rates, and temperatures were those specified. The net effect of this discrepancy would be to reduce the performance index of this unit. 5. The fan-blade angle for this test was less than the maximum attainable, due to mechanical difficulties. This reduction of blade angle would decrease the air velocity past the tubes. The reduced flow rate of air would not only result in.a lowering of the outside coefficient, but also cause a decrease in the mean temperature difference due to the greater temperature rise of the air. Since the performance index given as 65% was computed while assuming design flow rate and a heat balance, the real performance index should be appreciably higher than that indicated. 81

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN E. CONCLUSIONS lo Lean Oil Cooler.o-This unit cannot be expected to perform in accordance with the specifications. 2, Debutanizer Overhead Condenser. -This unit cannot be expected to perform in accordance with the specifications. 3. Evaporator Condenser.-This unit cannot be expected to perform in accordance with the specifications. 4. Still Overhead Partial Condenser.-This unit can reasonably be expected to perform in accordance with the specifications. 82

APPENDICES

- ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN - APPENDIX A CALCULATIONS ON DATA TAKEN IN LOVINGTON, NEW MEXICO CALCULATIONS ON THE PERFORMANCE OF THE JACKET WATER COOLER FROM DATA TAKEN THURSDAY, APRIL 28, 1955 A. East Bao — 1, Data. Velocity profile (see Table I) Average anemometer velocity = 539.9 ft/min Inlet air temperature (avg) = 72.1~F Outlet air temperature (avg) Ref. EMF Position (Mread (MV)32F Temp ~F 1.36 MV 3CR 1.875 350235 143.1 3CL 1.93 3.29 145.0 3DR 1.77 3.13 139.7 3DL 1.755 3.115 13901 3EL 1.82 3.18 141.4 3ER 1.715 3.075 137.9 avg = 141. ~F 4CR 1.46 2.90 131.7 1,44 4CL 1,625 3.065 136.4 4DR 1.50 2.94 133.0 4DL 1.715 3.15 140.4 4EL 1.50 2.94 133.0 4ER 1.49 2.93 132.7 avg = 134.5~F Average outlet temperature: 1410. + 134.5 2 = 137.8~F average temperature 83

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - Inlet water temperature (avg) = 162.5~F Outlet water temperature (avg) = 154.5~F 2. Calculation of heat duty. Q = W Cp At The face area of the tube bay is (allowing four inches for rolling in and end effects): Aby = (23567) x (92-5) x 2 bay ~~~~~12 366 ft2 Average indicated air velocity = 539.9 ft/min Average exit air temperature = 137.8~F The corrected face velocity (correction for density) = 550 539o9 597,8 508 std ft/min Standard density = 0.074 lb/cu ft Face area of the bay = 366 ft2 Lb/hr of air (indicated) = (60)(366)(0.074)(508) Air-side heat duty: = 825,000 lb/hr Q = Cp At Atair 137.8 - 72o1 = 65.7~F Cp(air) = 0.24 Btu/lb - ~F Q = (825,00ooo)(o24)(65.7) 13,000,000 Btu/hr 3. Calculation of ATM ATLM AT1 - AT2 AT1 AT2 b T2 AT1 = 154.5 - 72o.10 = 82.4~F AT2 = 162.5 - 1378 = 24.7~F

-ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - I ATLM = 82.4 - 24.7 n82,4 24.7 57.7 1.204 = 47.8~F ATLM correction (Kern) is negligible 4. Calculation of overall coefficient. The overall coefficient is obtained by: U = AATM From the Happy Co. specification sheets: AL = 2250 ft2 of liner area UL = 132000)(40 = 121 Btu/hr-~F-ft2 (liner area) (2250)(47.8) Uo = 121 - 8.84 Btu/hr-~F-ft2 (outside area) 1307 5. Calculation of corresponding water flow rate. Assuming an air-side —water-side heat balance, the water-side flow rate is: W Q Cp Atwater The water-flow area i 135, 000, 00 1,625,000 lb/hr (1)(162.5 - 154.5) Aflow = N Af low/tube Aflow/tube t (0 902)2/ e (4)(44) = o.o00443 ft2/tube (4) (144) Alow = (179)(0.00443) = 0.794 ft2 per pass per bay The water velocity inside the tubes is: V = 1,625,000 (0.794)(53600)(6) 9 t/sec 85

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN This is too high a velocity. 6. Water flow rate from the Happy Co. specification sheet. GPM = 3886 from specification sheet W = (3886) = 4.34 cu ft/sec/bay (2)7.48(6o) 4.34 V 494 = 547 ft/sec 0.794 The velocity from air-side data as computed in the previous section was 9.33 ft/sec. This value is 71* higher than specified. 7. Calculation of the outside coefficient. The outside coefficient will be computed from the relationship: 1 1 Ao Ao U0 = h r + rO + rm + A Aihi where Uo = overall coefficient of heat transfer, ho = outside film coefficient, ro = outside fouling film resistance, rm = metal resistance, ri = inside fouling film resistance, and hi = inside water film coefficient. a. Calculation of inside film coefficients The inside water film coefficient will be determined by hi 150(1 + 0.011 tw)Vt08 - 0.2 di (t W) 162.5 + 154.5 15 dwav =e ( 0 2 = O9 158.50F avg 2 di~2 = (0.902)'2 = 0O979 - 0.98.0 86

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - Using the water velocity computed from the air-side data, Vt.8= (933)08 = 6.o Substituting hi = 150[i + 0.011(158.5)]6.0 2520 hi 2520 0.98 Ao 3.59 x 12 A= = (t)(0.902 = 15.2 Ai (1r)(0.902 A0 _ 15.2 006 hr-~F.ft2 (outside area) Aih 2520 Btu b. Outside foulingO The outside fouling will be assumed equal to zero: ro = 0. c. Metal resistance. The computed outside coefficients will be compared with other corresponding coefficients which include in them a copper liner resistance and a root wall resistance; therefore, the metal resistance to be used here will be the difference between an admiralty and a copper wall resistance. The metal resistance will be given by XAo XAo rm AmKi AmK2 XAo K2K - K Am L K1K2' where X = liner metal thickness, ft, Ao = outside heat transfer area, ft2/ft, Am = mean metal heat transfer area, ft2/ft, K1 = thermal conductivity of admiralty, Btu/hr-~F-ft, K2 = thermal conductivity of copper, Btu/hr-~F-ft. x = 0- 49 o.oo408 ft, 12 87

LN ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Ao = 3559 ft2/ft, Am = 0.249 ft2/ft (liner), K2 = 220 Btu/hr-~F-ft (copper), and K1 = 64 Btu/hr-~F-ft (admiralty). Substituting (0.00408)(3.59) 220 64 rm 0- o249 (220)(64) = 0.000652 hr-"F-ft2 (outside area) Btu d. Inside fouling. Reference is made to Section VIII of this report, in which the inside fouling was determined to be riA = 0.0175 (based on the outside area). i Ai ~ e. Calculation of outside coefficient, 1 1 A0 A __ho rO, [ + Ari + r ho Uo iAih Aa Substituting from Sections a, b, c, and d: -1 18 -84 [0.o00604 + 0.000652 + 0.0175] h = 8~84 ho = 0.1151 - 0.02419 = 0.08s 8 ho = 11.26 Btu/hr- F-ft2 (outside area) ho('liner) = 11.26 x 13.7 = 154.5 (based on liner,area).. (The above values of ho are based on the field test data without the application of the duct correction factor. ) i 88

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN B. West Bay, — 1. Data. Velocity profile (see Table I) average anemometer velocity = 470.1 ft/mino Inlet air temperature (avg) = 75.7~F Outlet air temperature (avg): (Reference temperature = 27.8~C = 82.0~F = 144 Mv) Temperature ~F Position.,,. MV reading. 0C, M 1CR 1CL 1DR 1DL 1EL 1ER 1.45 1.60 1.465 1.485 1.49 1.47 2.87 5.04 2.905 2.925 2.95 2.91 150.6 136.4 151.7 13107 152.5 152.7 152.0 152.65~F Avg Position MV reading 52~F MV Temperature ~F 2CR 2CL 2DR 2DL 2EL 2ER 1.375 1.60 1.46 1.445 1.545 1.60 2.735 2.96 2.82 2.805 2.905 2.96 126.2 153.7 129.0 128.5 151.9 133 7 151.0 ~F Avg Average outlet temperature: 132.65 + 131.0 2 Inlet water temperature (avg) Outlet water temperature (avg) = 131.8~F = 163.5~F = 156,0~F 2. Calculation of heat duty. Q = W Cp At The face area of the tube bay is (see Section A) 366 sq ft. 89

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF M Average indicated air velocity = 470.1 ft/min Average exit air temperature = 131.8~F The corrected face velocity (correction for density) 470.1 445 ft/min. 591.8 Std density = 0.074 lb/cu ft lb/hr of air (indicated) = (60)(366)(0. 74) (445) = 723,000 lb/hr Cp = 0.24 Btu/lb-~F Atair = 131.8 - 75.7 = 56.1~F Q = (0.24) (56.l)(723,000) 9,7300,000 Btu/hr 3. Calculation of ATLM ATLM = AT AT AnTAT2 IICHIGAN - AT1 AT2 = 156.0 - 757 = =163.5 - 1318 = 80.3~F 31. 7'F 5235~F ATLM = 80.3 - 31.7 = 48.6 80n 3 0.93 31.7 ATLM correction (TEMA) is negligible. 4, Calculation of the overall coefficient. U = AATM From Section A-4 Substituting: UL AL = 2250 sq ft of liner area. 9,730,000 82 Btu (2250)(52.3) hr-OF-ft2(liner area) 90

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 82.8 Btu Uo = 62'6 605 ft2 =.1.7 = 6.5 hr-~F-ft2 (outside area). 5. Calculation of corresponding water flow rate. Assuming a water-side, air-side heat balance, the water-side flow rate is w = Q = 9,730, 000 Cp Atwater (1)(163.5 - 156) From Section A-5, the water flow area is Aflow = 0.794 sq ft per pass per bay. The water velocity inside the tubes is v - 1, 390,000 = ft./ (004 ( 3600) (61)t 390,000 lb/hr. - (0o.794)(5600)(6l) ".""' This is too high a velocity since the velocity computed from the specification sheet is 5.47 ft/sec (section A-6)o The velocity of 8.0 is 46% higher than that specified. 6. Calculation of the outside coefficient. The outside coefficient will be computed in the same manner as Section A. a. Calculation of inside film coefficient. 150(1 + 0.011 tw)Vt0'..2 di 0.2 di = (tw )avg 0.98 160~F Vt08 = 8,00 = 5.28. Substituting: 150[1 + 0.011(160)]5.28 0.98 2230 Ao Aihi 152 = 0.00681 (based on outside area). 2230 91

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN b Outside fouling. ro (assumed) = 0. c. Metal resistance (see Section A-7c). rm = 0.000652 (based on outside area). do Inside fouling (see Section A-7d). /Ao\ ri~( ) = 0.0175 (based on outside area). e. Calculation of outside coefficient. 1 1 h- = 6.o - [o.00oo681 ho 6. 050.1655 - 0.02496 ho = 1 = 7.13 0.1405 ho(liner) = (7.13)(1357) (The above values of ho are based on the tion of the duct correction factor.) C. Summary. — + 0.000652 + 0.0175] =.1405 (based on outside area) 97.7 (based on liner area). field test data without the applica Q, Btu/hr ATLM, ~F Uo (based on outside area) UL (based on liner area) Vt, ft/sec ho (ho)L Vface, std ft/min Water flow rate, lb/hr hwater flow East Bay 13 000,000 47.8 8.84 121 9.33 11.26 154.5 508 1,625,000 2520 West Bay 9, 730, 000 52 3 6.05 82,8 8.0 7.13 97~7 445 1,390,000 2230 Total 22,730,000 3,015,000 _rr 92

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - APPENDIX B DATA OF MONDAY, MAY 2, 1955 THREE-IN-LINE, FOURTH-ROW-STAGGERED BAY 1. Air-Side (Indicated) Heat Duty. —Average indicated air velocity = 785.6 ft/min Average exit air mev = 1.317 Average cold junction temp = 29.1~C = 84.4~F 84.4~F = 1.50 mev Total EMF = 1.50 + 1.317 = 2.817 Corresponding outlet temperature = 128.5~F Temperature rise of air = 128.5 - 80.8 = 47.7~F Lb per hour of air = 785.6 ft/min at 128.5~F and 29.87 in. Hg Corrected face velocity = 785.6 30/588.5 = 747 std ft/min Std density = 0.074 lb/cu ft Face area = 366 sq ft (see Appendix I, Section A) Lb of air per hour (indicated) = (366)(747)(o.074)(60) = 1,215,000 lb/hr Q = (1,215,000)(024)(47.7) = 13,900,000 Btu/hr. 2. Heat Transferred From Water-Side Data.-For comparison purposes, the heat duties computed from the water-side measurements are as follows: Inlet water temperature = 152.7~F Exit water temperature = 145o6~F. water temperature drop = 910~F average water temp = 148~F average water density = 61.3 lb/cu ft The water velocity was determined by an orifice-plate pressure-drop measurement. The orifice diameter was 11.25 in. The water pipe was a 16-in. schedule-20 pipe (ID = 153.75 in. = 1.286 sq ft). The manometer reading was 66 in. of water (5.5 ft). The relationship for determining the velocity of 93

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN flow through the orifice is given by* V - C\ 2g Hv q SNo. (~ ) si2 where Vo = orifice flow velocity, ft/sec, SO = orifice flow area, sq ft, C = orifice coefficient, dimensionless g = acceleration of gravity, 32.2 ft/sec2 Hv = pressure drop across the orifice, feet of flowing fluid, and Si = cross-sectional flow area of the pipe, sq ft. By a trial-and-error procedure the orifice coefficient is determined from a Reynolds number plot.** The Reynolds number is over two million, therefore the orifice coefficient is taken equal to 0.61. The value of Hv is 5o5 ft. Therefore, Vo - 0 ~6lM2(32e2)( e ) Vo o.6 -,) 1 - (o0555)2 0.61(18.8) o.845 = 13. 58 ft/sec. The pipe velocity is V = 13.58(0o535) = 7.26 ft/sec. The lb per hour of water flowing through the heat exchanger is 7.26(l.286)(3600)(6103) = 2,060,000. It is assumed that since this water was split between two units, one-half of the water flowed through the unit under consideration. Therefore, the total heat transferred is Q = (w Cp At)water 2,06,000 (1)(9.1) 2 = 9,380,000 Btu/hr. *G. G, Brown et al., Unit Operations. New York: John Wiley and Sons, 1950, p. 158. *Loc. cit. 94

ENGINEERING RESEARCH INSTITUTE * U 3~ Comparison of Heat Duties — Air side = 13,900,000 Btu/hr Water side = 9,380,000 Btu/hr Discrepancy = 4,520,000 Btu/hr NIVERSITY OF MICHIGAN % greater than water side = 4, 20, 000\oo \9,380,00ooo = 48% 4. Calculation of Overall Coefficient From Air-.Side Data.U = AATLM ATLM AT -, AT2 AT1 AT2.AT1 143.6 - 80.8 = 62.8~F AT2 = 1527 -128.5 = 24.2~F ATL 62.8 - 24.2 62.8 Qn 24.2 38.6 0.954 40.5~F Aliner = 2250 sq ft Substituting: UL UO 13,900,000 (2250)(40.5) 152.8 (based on liner area) =- 12.8 15.7 = 11.15 (based on outside area). 5. Calculation of Corresponding Water-Side Flow Rate? Q = C' At 15,900,000 - (1)(152.7 = 1,528,000 lb/hr - 145.6) Water flow area: Aflow = 0.794 sq ft per pass per bay. The water velocity inside the tubes is 95

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Vt 1(528,)000 = 8.76 ft/sec. t (0.794) (61) (3600) 6. Calculation of Air-Side Coefficient From Air-Side Data. —The procedure is the same as in Appendix A. a. Calculation of water film coefficient. 150[1 + 0.011(147.7)](8.76)0'8 hi o.g8 0.98 2270 Ao 152 0.0067. Aihi 2270 b. Outside fouling. ro = 0 (assumed). c. Metal resistance (see Section I-A7c). rm = 0.000652 (based on outside area). d. Calculation of outside coefficient, assuming no fouling. 1 1 -- =.1 -1 [0o0067 + 0o000652] h o 1.15 = 0o0896 - 0.00735 = 0.08225 h = _-1 — = 12.18 (based on outside area) 0 0.o08225 (ho)L = 12.18 x 13.7 = 166.5 (based on liner area). e. Calculation of outside coefficient, assuming inside fouling of 0.001 (based on inside area). 1 = 0.0896 - [0,00735 + 0.0152] h0 1 h 0.6705 = 14.l95 (based on (hL = ( 95)(17) = 205 (base (ho)L = (14.95) (13.7) 205 (base( = 0.06705 outside area) d on liner area). 96

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 7. Calculation of Overall Coefficient From Water-Side Data and Measured Air Temperatures. u = _Q A ATLM where Q = the water side heat duty of 9,580,000 Btu/hr, A = 2250 sq ft of liner area, and ATLM = 40.5~F (see Section II-A4). Substituting: 9,80,000 _ UL = (2250)(40.5) = 103.0 (based on liner area) U ~ = 103.0 7.52 (based on outside area). 8. Calculation of Corresponding Air-Side Flow Rate. W' = Q - 9,380,000 820,000 lb/hr. Cp At (0.24)(1285 - 80.8) The corresponding face velocity is 820,000 Vface = (66(60)(0074) = 504 std ft/min. 9. Calculation of Air-Side Coefficient From Water-Side Data — a. Calculation of water film coefficient., 9,580,000 0.8 hi = 2270 (l5,900,00o) = 1655 (based on inside area) Ao 15.2 Ah = 1655 = 0.0092 (based on outside area). Aihi 1655 b. Outside fouling assumed zero. c. Metal resistance (see Section I-A7c). rm = 0.000652 (based on outside area). d. Calculation of outside coefficient, assuming no fouling. 1 1 7h52 = - -[0.0092 + 0.000652] = 0.1241 ho 7.52 97

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - ho = 01241 = 8.05 (based on outside area) ho= 0.1241 (ho)L = (8o05)(1357) = 110 (based on liner area). e. Calculation of outside coefficient, assuming an inside fouling of 0.001 (based on the inside area). h- = 7.52- [0.0092 + 0.000652 + 0.0152] = 0.1089 ho 7.52 ho = 1 = 9.18 (based on outside area) 0.1089 (ho)L = 9.18 x 13.7 = 126 (based on liner area). 98

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN APPENDIX C TRIANGULAR-PITCH BAY 1. Air Side (Indicated Heat Duty). Average indicated air velocity = 661.8 ft/min Average- exit air mev = 1.0335 Average cold-junction temperature = 35.03~C = 95.1~F 95.1~F = 1.8228 my Total EMF = 1.8228 + 1.0335 = 2.8563 my Corresponding outlet air temperature = 130.2~F Average air inlet temperature = 79.5~F Temperature rise of air = 130.2 - 79.5 = 50.7~F Lb per hour of air: 661.8 ft/min at 130.2~F and 29.87 in. Hg Corrected face velocity = 661.8 530/590.2 = 626 std ft/min Std density = 0.074 lb/cu ft Face area = 366 sq ft Lb of air per hour (indicated) = (366)(626)(0.074)(60) = 1,019,000 lb/hr Q = WCp At = (1, 019,000)(0.24)(50.7) = 12,410,000 Btu/hr. 2. Heat Transferred From Water Side. Lb of water Temperature Temperature per hour (from previous section) of inlet water = 148.7~F of outlet water = 140.5~F At of water = 8.2~F = 1,030,000 Q = (1,030,000)(1)(8.2) = 8,450,000 Btu/hr. 3. Comparison of Heat Duties.Air side = 12,410,000 Btu/hr Water side = 8 450,000 Btu/hr Discrepancy = 3,960,000 Btu/hr Percent greater than water side = 32960000 = 46.9%. 8,450,000 99

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 4. Calculation of Overall Coefficient From Air-Side Data. — U = Q AATLM TL _ AAT1 - AT2 LM- AT1 AT2 ATI = 140#i5 - 79.5 61.0~F ATS = 148.7 130.2 = 18.5~F. Substituting: AT 61.o - 18...5 TLM In 61.0 18.5 12,410,000 UL (2250)(34.7) 159 41. o _ 34.7~F 1.195 (based on liner area) 1597. Calcula15 11.62 (based on outside area),5 o - 3.7.5. Calculation of Corresponding Water-Side Flow Rate. W 12,410,000 (1.0)(148.7 - 140.5) v = 8 e76 5l7,000\= Vt = 876,528000 - = 1,517,000 lb/hr 8.84 ft/sec. 6. Calculation of Air-Side Coefficient From Air-Side Data. a. Calculation of water film coefficient. 0.8 (150)[1 + 0.011 (145.1)](8.84) hi = 0.98 hi = 2280 100

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Ao _ 15 2 _ 0.0067 Aihi 2280 bo Outside fouling assumed zero. c. Metal resistance (see Section I-A7c). rm = 0.00062 (based on outside area). d. Calculation of outside coefficient, assuming no fouling. 1 = - [0.0067 + 0.000652] = 0.0786 ho 11.62 1 ho = 0 0786 = 12.7 (based on outside area) (ho)L = 12.7 x 13.7 = 174 (based on liner area). e. Calculation of outside coefficient, assuming inside fouling of 0.001 (based on inside area). 1= 0.0860 - [0.007352 + 0.0152] = 0.06345 ho h = 106 = 15.8 (based on outside area) 0 0Q6.0345 (ho)L = 15.8(13.7) = 216 (based on liner area). 7. Calculation of Overall Coefficient From Water-Side Data.Q u AAT ATL = 34 7~F A = 2250 sq ft of liner area Q = 8,450,000 Btu/hr (based on water-side data). Substituting: 8,450,000 = 108 (based on liner area),L (2250)(34.7) UO= 18 = 7.90 (based on outside area). 13.7 101

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 8. Calculation of Corresponding Air-Side Flow Rate. A. w Q... 884. Cp At ir (0.24) (13C The corresponding face velocity is 50, 000 = 695,000 lb/hr. ).2 - 79.5) v 9,,. 428 std ft/min. face (366)(60)(0.074) 9. Calculation of Air-Side Coefficient From Air-Side Data. - a. Calculation of water film coefficient. h = 22808,450,0o hi= 2280 2,basdooo/ 167 (2410,based on inside 00area) 1673 (based on inside area) A0 = 15.2 =0.00908 (based on outside area). Aihi 1673 b. Outside fouling assumed zero. c. Metal resistance (see Appendix AI-7c) rm = 0.000652 (based on outside area). d. Calculation of outside coefficient, assuming no fouling. 1_ = 1 9~ [0.00908 + 0.000652] ho 7.90 = 0.1266- 0.00972 = 0.11687 1 ho = o.1687 8.56 (based on outside area) (ho)L = 8.56 x 13.7 = 117 (based on liner area). e. Calculation of outside coefficient, assuming inside fouling of 0.001 (based on inside area).:,! = 0.11687 - 0.0152 = 0.1015 ho 102

-- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1 ho = 0.1015 = 9.85 (based on outside area) (ho)L = (9.85)(13.7) = 135 (based on liner area). 103

-- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN APPENDIX D CALCULATION OF THE PERFORMANCE OF THE JACKET WATER COOLER, USING AN ANEMOMETER DUCT FACTOR OF 0.636 A, DATA OF THURSDAY, APRIL 28, 1955 1. East Bay. a. Heat duty. The indicated face velocity (see Appendix A) of 508 std ft/min is to be multiplied by 0.636 (see Fig. 2 of this report). Vface = 508(0 636) = 323 std ft/min Aface = 366 sq ft of face area Pstd = 0o074 lb/cu ft W = VpA = (323)(366)(0.074)(60) = 525,000 lb air/hr Q = W Cp At At = 137.8 - 72.1 = 65.7~F (see Appendix A) Cp = 0,24 Btu/lb-~F Q = 525,000(65.7)(0 24) = 8,290,000 Btu/hr. b. Log-mean temperature difference. ATLM 47.8~F (see Appendix A) c. Calculation of overall coefficient. The overall coefficient is given by: 104

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN U = Q A ATLM AL = 2250 sq ft of liner area = 8,290,000 - 77.0 (based on liner area) (2250)(47.8) UL UO = -3 = 5.61 (based on outside area). do Calculation of corresponding water flow rate. From Appendix A: VT = 9533 x 0.636 = 5.95 ft/sec. eo Calculation of the outside coefficient. (1) Inside (water) film coefficient. 0.8 hi = 2520 = 1766 \9.533 A0o 15 2 = 0.00862. Ai hi 1766 (2) Outside fouling assumed zero. (3) Metal resistance (see Appendix A), rm = 0.000652 (based on outside area). (4) Inside fouling (see Appendix A). Ai ri = 0.0175 (based on outside area). (5) Outside air film coefficient (5) Outside air film coefficient. 1 AO Ao h h = U -o - i + Airi + rm + rl ho UO Ai hi Ai 0 105

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Substituting from (1), (2), (3), and (4): 1 = 1- - [0.00862 + 0.0175 + 0.000652] ho 5.61 = 0.178 - 0.02677 = 0.1512 ho = 6.63 (based on outside area) 0.1512 (ho)L = (6.63) (1357) = 90.8 (based on liner area). 2. West Bay,a. Heat duty. Using the duct factor of 0.636: Vface = 445 x 0.636 = 283 std ft/min W = 723,000 x 0.636 = 460,000 lb/hr lb of air per hour (see Appendix A) Q = 9,730,000 x 0.636 = 6,180,000 Btu/hr. b. Log-mean temperature difference (see Appendix A). TLM = 52.3~F. c. Calculation of the overall coefficient. UL A ATLM (225)(523)7 (b d on liner area) Uo = 53 = 5.85 (based on outside area). 13.7 d. Calculation of the corresponding water flow rate (from Appendix A). Vt = (0.636)(8.o) = 5.09 ft/sec. t 106

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN e. Calculation of the outside coefficient. (1) Inside water film coefficient. 0.8 hi = 2250 (i09) = 1552 (based on inside area) A0 - 15.2 = 0.0098 (based on outside area). Ai hi 1552 (2) The outside fouling will be assumed zero. (3) Metal resistance. rm = 0.000652 (based on outside area). (4) Inside fouling (see Appendix A). A O A ri = 0.0175 (based on outside area). A(5) Outside air film coefficient. 1 1A A0 - = — o L i + A ri + rm + ro ho Uo Ai hi Ai Substituting: 1 1 - == - - [0o0098 + 0.0175 + 0.000652] ho 3.85 h = 0.260 - 0.02795 = 0.2315 4.2 (based on outside area) o - 4.32 (based on outside area) 0.232 (ho)L = (4.32)(13.7) = 59.1 (based on liner area). B. DATA OF MONDAY, MAY 2, 1955 1. Three-in-Line, Fourth-Row-Staggered-Pitch Bay..107

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN a. Heat duty. The heat duty calculated from the air-side data is directly proportional to the air velocity. Therefore, Q = (Q) indicated x:0.636 = 13,900,000 x 0.636 = 8,840,000 Btu/hr. The discrepancy between the computed air-side heat duty and the computed water-side heat duty then is 9,380,000 - 8,840,000 = 540,000 Btu/hr, which is 5,76f% of the water-side heat duty. The face velocity is Vface = (747)(06536) = 475 std ft/min. b. Log-mean temperature difference (see Appendix A). ATM = 40.5~F. c. Calculation of the overall coefficient from the air-side performance data. U = Q A ATLM Substituting from a and b, with a liner area of 2250 sq ft: UL' = 8,840,000 UL = (2824005) = 97.0 (based on liner area) 97_ o UO = = 7.08 (based on outside area). d. Calculation of the corresponding water flow rate. Vt = (8.76)(0.636) = 5.58 ft/sec. e. Calculation of the outside coefficient. (1) Inside water film coefficient (see Appendix B-6). hi = 2270 (8.76) = 1580 (based on inside area) 8 <76 108

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Ao 15.2 Ai hi 158 = 096 (2) The outside fouling will be assumed zero. (5) Metal resistance. rm = 0,000652 (based on outside area). (4) Calculation of ho, assuming no fouling. 1 1 h- = 7.08' [.09oo6 + 0,000652] = o.1411 - 0.01025 = 0,15085 o 01308 7.64 (based on outside area) 0 0.13085 (ho)L = (7.64)(13.7) = 105 (based on liner area). (5) Calculation of ho assuming an inside fouling factor of 0.001 (based on the inside area). 1 = 0 085 - 0152 = 0.18 0012 1565 ho 1 ho = O 11565 = 8.65 (based on the outside area) (ho)L = (8.65)(1357) = 118.5 (based on the liner area). 2. Triangular-Pitch Bay. — a, Heat duty (see Appendix C). Q indicated (air side) = 12,410,000 Btu/hr Q = Q indicated x 0.636 = 12,410,000 x 0.636 - 7,900,000 Btu/hr. The discrepancy between the computed air-side heat duty and the computed water-side heat duty is 8,450,000 - 7,900,000, or 550,000 Btu/hr, which is 6.50% of the water-side heat duty. The corrected face velocity is 109

[ ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN - Vf = (626)(0.636) = 398 std ft/min. b. Log-mean temperature difference (see Appendix C). ATLM = 34.70F. c. Calculation of the overall coefficient from the air-side performance data. U = - A ATLM 7,900,000 UL - (2250) (54.7) 1= 101.3 (based on liner area) 101o3 O. U = 113. == 7.40 (based on outside area). I d. Calculation of the corresponding water-side flow rate. Vt = (8.84)(0.636) = 5.63 ft/sec. e. Calculation of the outside coefficient. (1) Inside water film coefficient (see Appendix C). hi = 2280 (8 0 hi = 1590 (based on inside area) Ao 15.2 =-oo0056 Ai hi 1590 (2) The outside fouling is assumed zero. (5) Metal resistance. rm = 0.000652 (based on outside area). (4) Calculation of outside coefficient, assuming no fouling. h- = I - [0.o00956 + 0.000652] = 0.155 - 0.0102 hO ^ 7.4 i 1 h= 0.1248 = 8.02 (based on outside area) 0,Ol4 110

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN (ho)L = (8.02) (1.7) = 109.8 (based on liner area). (5) Calculation of outside coefficient, assuming an inside fouling factor of 0.001 (based on inside area). 1 - = 0.1248 - 0.0152 = 0.1096 h ho = 9.13 (based on outside area) (ho)L = (9-13)(13.7) = 125 (based on liner area). 111

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN APPENDIX E EVALUATION OF AIR-SIDE COEFFICIENT BY 651C FOR THREE-IN-LINE, FOURTH-ROWSTAGGERED UNIT A calculation will be presented for a one-in.-OD-liner tube containing 9 fins per inch, having a nominal two-in. diameter over the fin. 13. Btu/hr, of 200~F Reference is made to 651C, pages 8 and 9, Tables 10, 11, 12, and According to Table 12, the tube length required to transfer 10,000 using this tube under a log-mean temperature difference driving force with a maximum air velocity of 1457 ft/min, is 10.9 ft. According to Table 11, 3000 cu ft/min of air passing through an air-duct cross section of five sq ft results in a maximum air velocity of 1457 ft/min. The corresponding face velocity is therefore V face 3000 600 ft/min. face 5 The overall heat transfer coefficient is given by Q U Ao AT According to Table 10, this tube has an Ao of 3.61 sq ft/ft and an Ao/Ai of 14.8. Substituting: 10,000 U = (10.9)(3.61)(20) = 12.7 Assuming an industrial steam coefficient of 1000, with no fouling, and neglecting the metal resistance, the overall coefficient is given by 112

- ENGINEERING RESEARCH INSTITUTE * I 1 Uo = - Ao ho Aihi JNIVERSITY OF MICHIGAN - or 1 1 Ao Uo Aihi 1 1 0.0639 = 157 1 w 12.7 14.8 1000 The ratio of Vmax is equal to 1457 = 2.428 Vface 600 The ho, if expressed on liner area, becomes (h)liner = 15.7 x 0 Oliner - 04262 = 216.5

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN APPENDIX F CALCULATION OF TRIANGULAR AIR-SIDE COEFFICIENTS AND PRESSURE DROP FROM THE CORRELATION REPORT, USING MONDAY'S TEST DATA From Appendix C the face velocity measured was Vface 398 std ft/min. The corresponding maximum air velocity is Vmax = 2.428 x Vface = 2.428 x 398 968 ft/min. The dimensions of the tube tested, as given in the Wolverine Trufin Catalog "C,. are: Catalog Number 62-0916049-01, 1-in.-OD liner 18 B.W.G. tube. OD fins: 1.938-2.063 in. mean fin thickness: 0.019 in. outside area: 3.59 sq ft/ft ratio of outside to inside area Ao/Ai: 15.28. 1. Calculation of AirSide Coefficient, Using Fig. 1. —From Fig. 135 at a maximum air velocity of 968 ft/min, /ho S\ 0.5 Dr)=(NDo) 89, where the terms included are defined in Report No. 30, page 51. S =average tube pitch from blue print -1/16 + 2-1/8) 20094 in., Dr = 1.08 in. (assuming a 0.04-ino root wall metal thickness)., N = 9 fins per inch, and Do = 2.00 in. 114

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Substituting: h o 2.094 0.5 2-094. [(9)(2)] - 89 ho' = 10.85 based on the equivalent outside area. The outside coefficient given is this report has to be corrected for the fin efficiency, using the relationship ho' Aeq [ho]outside Ao, where Aeq = Ar + Ef Af, Ar = area of root, sq ft/ft, Af = area of fin, sq ft/ft, and Ef = fin efficiency. The root area of this tube is Ar = Dr (1 - NY), 12 where Y is the fin thickness. Substituting: Ar = 1 (1.08)[1 - (9)(0.019)] A12 = 0.232' 0.23 sq ft/ft Af = 3.59 - 0.23 = 3.36 sq ft/fto Determination of fin efficiency: From the attached plot on the next page, the abscissa is abscissa = H 2H, ( + r KmY +o (.v o)V " since the foulng on the outside of the tube is assumed to be zero. Substituting the values of ho' = 10.85, 115

- ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN — 7 1.0 0.9 0.8 07.0.7 I I I I IV -I. I i i i. i i i — - -- - -I - - L I Iy d~ Y/2 I D |V l-azizzfez 11 1 \ n \ThZ7 \ I6 LiJ 0 z L k. IL z U. A a N il \ 0. —----- 0.2 ----- 0.2 —... C). l — - - - - - - - - __ _ _ - -. W w I.. II I 0 0 1.0 H',.)fKY 2.0 3.0 Fig. 25. Gardner's fin efficiency. 116

-- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN H = 2.0 - 1.o8 = 0.0384 ft, (12)(2) c n = y = abscissa = 119 Btu/hr-~F-ft 0.019/12 = 0.00158 ft, 0584 (2)(10.85) 0.0384 (119) (0.00158) (ig( o 18 = o0413. The parameter, Do/Dr, 2 -00 1.85. 1.08 From the figure, Ef = 0.905 = 90.5o. The equivalent area is Aeq = Ar + EfAf = (0.25 ) + (0.905)(3.36) 3.27 sq ft/ft. The outside coefficient then becomes [ho]outside = ho,( ) = A"O 10.85 (3 ) = 9.89, or (ho)L based on liner area = 135.5. 2. Calculation of Air-Side Coefficient, Using Fig. 14.-The Reynolds number for finned tubes is defined as De Gmax Re = where 2 De - (- "Y f Gmax = maximum mass rate of flow through the tube bank, lb/hr-sq ft, and,. = average viscosity of fluid, lb/ft-hr. Using the dimensions given in the previous section, De = () (12) = (\1 8 J 12 ft 117

[ | - ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Gmax = Vmax p = (60)(968) x 0.074 = 4300 lb/hr-sq ft |i* = o0.46 lb/ft-hr at 100~F. Substituting, Substituting, De Gmax (5,62)(4300) - ^Re 5 0o - 525,000. From Fig. 14, at Re = 525,000, -Oo 33 ho' De (C ) = 4700 K K +0 * 33 K +0.33 = (0.71) 0.892 K = 0.0157 Btu/hr-~F-ft. Substituting these values, h^t (4700)(0.0157) (0.892) — v 5.62 11.70, based on the equivalent area. Following the same procedure as given in the previous section, the abscissa of the fin-efficiency plot is (2)(11,70) abscissa = 0.-0384 )(1) (119)(0.oo158) = 0,428. From the fin-efficiency figure, Ef = 0,90. Substituting, Aeq = 0.23 + (0.90)(3.36) = 3.25 sq ft/ft, the outside coefficient, corrected for fin efficiency is *The physical properties of air are obtained from: W. Ho McAdams, Heat Transmission, third edition. New York: McGraw-Hill Book Co., 1954, p. 483. - 118

- ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN - (ho)outside = 11.7 (3 ) = 10.6 (based on outside area) 3.59/ (ho)liner = 13.7(10.6) = 1450 53 Calculation of Air-Side Coefficient, Using Fig. 15. —From this figure at a maximum air velocity of 968 ft/min, 0.3 (hoS) ( = 36 = 1.89 N 0.3 (Dr - 9 0~3 1.08) S = 2o094 in. (avg). Substituting: 36 (2.094) (1.89) = 9.11 (based on equivalent area). The abscissa of the fin-efficiency plot is abscissa = 0.05844 (2)(9.11) 0.378. (119) (.oo00158) From the fin-efficiency figure, Ef = 0.93, Aeq = 0.23 + (0,93)(3536) 3.35 sq ft/ft. The outside coefficient corrected for fin efficiency is ho 9- 11 335 = 8.5 (based on outside area) (ho )liner = ( (8 59o6 (ho)liner = (13.7)(8.5) - 116,4~ 4, Pressure Drop, Using Fig, 16.-The parameter given on this figure for this tube is N Do K = 0.2 Dr 9 x 2.0 (1 08)0 = 17.75. 119

'- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - For a maximum air velocity of 968 ft/min, the pressure drop times the tube equivalent diameter is -AP ~ De = 0.50 ft (inches of water per row) De = 5.62 ft -AP = 0.089 in, of water per row. The pressure drop for a bank four rows drop is (4)(.089) = 0.356 in. of water. 120

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN APPENDIX G SCHMIDT IS CORRELATION Reference is made to Fig. 19 in which the Schmidt group is plotted as a function of maximum air velocity. The maximum air velocity of 968 ft/min gives a value of 15 for the Schmidt group. Fho D ~04 1 ho r_,,-1 = 15. r r. T) The dimensions of the tube (see Appendix F) are Do = 2.00 in., Dr = 1.08 in., and N = 9 fins/per inch. Therefore, Dr04 Dr = (1.08)0~4 = 1.032 DO - Dr 2 2.00 - 1.o8 = 0.46 in. 2 [EN ( Do Dr j\O0.63'Do 2 —- f.4063 = [9 x o0.46] 2.445. Therefore, D0.4 D r 1.032 1.032 0.56 [1.D- 0.8- Dr)\ N 63 L~ ~~~- O~ 8[( 0 1 1 - 0.18(2.445) 121

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Therefore, ho (1)(.6) = 8.14 ~0 1.052 The abscissa of the fin-efficiency plot is 2ho= (2) (814) abscissa = 00384 (2)(8 o057 2KmY (zl9)(o.oo158) Do 2.0.. 1.850 Dr io8 From the figure, Ef = 0o945. Using the area values given in Appendix F, Aeq = Ar + EfAf = 0.23 + 0o945(3.36) = 3.41 sq ft/fto The outside coefficient corrected for fin efficiency is ho = ho' (A ) = 8.14 (35) = 7.73 (based on outside area) (ho)L (7 (17) (based on lner area) ('^^ ^ (T"T5)(L5.7) = 106 (based on liner area)o L 122

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN APPENDIX H SAMPLE CALCULATIONS FOR FIG. 20, "ho vs HP/ft2" A. STAGGERED PITCH Catalog tube No. 62-0716065-01 Distance from center to center: No. fins per inch: N = 7 Ao Dr = 1.08 in. Vmax/Vface Face area: 366 sq ft Assume Vface = 600 std ft/min S = 2-1/16" = 2.0625 in. = 2.83 sq ft/ft Do = 2.0 in. = 2.34 0 Vmax = 1404 std ft/min 1. Calculation of ho. —From Fig. 13: (entering with Vmax of 1404) " (Dr No) 110 S (D)5 Dr 2.0625 1.08 (7 x = 7.15 * h 110 -ho7.1 7.15 ho' Btu sq ft (outside equivalent a15.4rea) hr-~F-sq ft (outside equivalent area)' This value must be referred to actual outside area. hot Aoeq = hoAo, or ho = ho Aoeq Ao Aoeq = Ar + 0 Af 3.14 x.1.08 Ar =.283 sq ft/ft 12 123

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Af = 2.83 - 0283 = 2.547 sq ft/ft. Fin efficiency is determined, using Gardner's plot: 2 abscissa: H (1 + ro)KmY H = 092 = 0o0383 ft 24 Km = 120 Y = 019 12 ro 0 abscissa = 0.03835\ 20 x 2 19 To this abscissa corresponds a fin efficiency of 0 =.895 = 487. AOeq 0.283 + (0.895)(2.547) = 2.563 Aoe - ho = ho' A Ao 2. Calculation a. Pressure drop From Fig. 16: K = 154 2.565 95 Btu 2.830 1.95 hr-~F-sq ft (outside area). of HP/sq ft Face Area.NDo - 7 x 2 (=02 1.08)0.213.8 Dr 0.2 (1o0 8 )0.2 _ /S2 ND (_ 2.0625 7 x 2 _ 428 t e ( Dr) (12 (1080) 12 ) 4.28 ft. For a value of Vmax = 1404 std ft/min, Fig. 16 gives AP De = 0.7 (in. water per row)ft 124

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - 3.4 lb/sq ft APt 0.7 x 4 x 62.4 12 x 4.28 b. Theoretical power input per sq ft face. P f=ft lb (3.3 x min 10-5) i x 105) HP/ft2 = (Vf) x (Apt) (3-03 600 x 3.4 x 3.03 105 = o0618 ft2(face area) Figure 20 is obtained by plotting for each assumed Vface, ho vs HP/ft2 (face area). B. THREE-IN-LINE, FOURTH-ROW-STAGGERED ARRANGEMENT This curve has been constructed by using the ALCO data given in Fig. 11o 1. For an Assumed Value of Vface, Read the Corresponding Value of ho From ALCO Line o Vface = 600 std ft/min 2. From the Same Plot, AP =... -... ho = 9.2.37 in. of Water for Four Rows — 4 1.92 lb/sq ft AP.370 x 62J, 12 Theoretical HP ft2 1.92 x 600 x 3003 =.0349 HP ft2 (face area) The corresponding graph can be obtained by assumed value of Vfaceo plotting ho vs HP/ft2 for each 125

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN APPENDIX I CALCULATIONS FOR LEAN OIL COOLER (ITEM 2-5) A. DESIGN CONDITIONS FROM SPECIFICATIONS SHEET (See Section XI of this report). 1. Air Side.Inlet air temperature: 100~F Air flow rate: 46,500 st cu ft/min Temperature rise of air: 30.5~F. 2. Tube Side,.Inlet oil temperature: 173~F Outlet oil temperature: 140~F Average molecular weight: 200 lb/mole Oil flow rate: 90,659 lb/hr Mean temperature difference: 41.2~F Number of passes: 10 Heat exchanged: 1,538,000 Btu/hr. 3. Heat Transfer Surface.Bare tube: 376 sq ft External area: Ao = 5180 sq ft Tube specifications: 1 in. OD x 18 B.W.G. Inhibited Admiralty Tube inside diameter: 0~902 in. Fin height: 1/2 in. nominal, 9 fins per inch Fin thickness: 0.019 in., Ao = 3.59 sq ft/ft, Ao/Ai = 15.28. B. PHYSICAL PROPERTIES OF LEAN OIL Specific heat: 0.520 Btu/lb-~F (Maxwell's Data Book on Hydrocarbons, p. 93) Specific gravity at 60~F: 0.825 (Given in specification sheet) Viscosity at 156.5~F: 1.48 cps (Maxwell's Data Book on Hydrocarbons, p. 161) 126

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Thermal conductivity at 156.5~F: 0.0780 (Btu/ft)/sq ft/hr-~F (Kern's Process Heat Transfer) API = 390 C. SPECIFIED HEAT LOAD Qoil = W Cp toil = W Cp Atoi1 Cp = 0.52 Btu/lb-~F Atoil = 173 -140 = 33 F. Substituting: Qdesign = (90,659)(0.52)(33) = 1,552,000 Btu/hr. This value compares with the heat duty given in the specification sheet as 1,538,000 Btu/hr. D. CALCULATION OF EXIT TEMPERATURE OF THE EXISTING UNIT, USING THE ALCO DATA CONVERTED TO 9 FINS PER INCH 1. Air-Face Velocity. —From the blue print supplied by The Happy Co., the width of unit is 93-5/8 in. The tube length is given in the specifications as 8 ft. Allowing one-in. stripped ends for rolling in, the face area is Aface = (9'25(7.832) = 61.1 sq ft. The corresponding specified air velocity (face) is then 46 50oo Vface = 61.1 760 std ft/min. 2. Outside Film Coefficient. — From Fig. 11, using the dotted line corresponding to 9 fins per inch at a Vface of 760 std ft/min, ho = 9.7 (based on outside area) (ho)L = (907)(137) = 133 (based on liner area). 3. Metal Resistance.-The metal resistance is obtained from Section I of Appendix A as 127

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY C rm = 0.00062 (based on outside area). )F MICHIGAN - 4. tained, using Inside Film Coefficient.-The inside coefficient will be obthe following relationship given by McAdams*: 2/3 h /Cp \/ C G\ K/ II c,14 0.023 DG)0~2 )/ where the symbols are defined by the author. The area of flow for the oil is Af low (18 tubes/pass) 0.902)2 (18 tubes/pass) (0902) (4)(144) 0.0798 sq ft G W Af low 90,659 0.0798 1,156,000 lb/hr-sq ft. DG 0*2 ie0.2 = 1.48 centipoise x 2.42 (0.902)(1,136,o000oo 0 3.58 lb/ft-hr r[2p.8001 L (12) (3.58) J.-. —) -/ ) VV J 5.9)2/3 - (2ul 9) = 751 8.3. 2/3 (o.052)(3.58)2/ 3 0.078 Substituting: (o.023) (0.52)(1,136, oo0) (7.51) (8.3) 0.14 IPMW/ 0.14 = 218 (I Assume that (14 is 099 (to be checked later Assume that (JL/|Jw) is 0.99 (to be checked later): hi = (218)(0.99) 216 (based on inside area). 5. The Inside Fouling is Obtained From TEMA as 0o002. —** *Op. cit., eq. 9-1Oc, p. 219. **Standards of Tubular Exchanger Manufacturer's Association, third edition, 1952, p. 87, Table T-2. 128

ENGINEERING R'ESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 6. The Overall Heat Transfer Coefficient is Given by.1 1 AO Ao = - + rm + - ri +Ah Uo ho Ai Ai hi Substituting: U1 1= 1 + 0.00062 + (15.28)(0.002) + 1528 Uo 9-7 216 0.105 + 0o00062 + 0.05056 + 0.0708 = 0o20498 UO = 020 = 4.88 (based on outside area) 0.205 (Uo)L = (4.88)(13.7) = 66.9 (based on liner area). 7. Calculation of Heat Duty. —Trial and error for the outlet temperature: assume toil out = 147.5 ~F Q = (90,659)(0.52)(173 - 147-5) = 1,210,000 Btu/hr. The temperature rise of the air is Q Atair W Cp 1,200,000 24 2~F (46,5oo 0)(6o0)(0.0o74)(o.24) - The outlet air temperature is then tair out = 100 + 24.2 = 124.2~F. The mean temperature difference driving force therefore can be obtained as AT1 = (173 - 124.2) = 48.8~F AT2 (147.5 - 100) = 47.5~F ATm = ATaverage = 48.1~F. Check on heat duty: Q = Uo AoATm = (4.88)(518o)(48.l) = 1,215,000 Btu/hr, 129

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN which is reasonably close to the value of 1,200,000 Btu/hr obtained by assuming an outlet oil temperature of 147.5~F, 8. Check of Inside Coefficient.-The temperature drop across the oil film is obtained from Atoil Uo AT^ h.^_ mT, ~ Ai jihi) Ao Y Substituting: (4.88)(15.28) (48.1) = 16.6F Atoil (48.1) (216) t(oil at wall) = 173 1475 _ 16.6 = 143.6~F 2 Lw = 1.58 centipoise (at 143.6~F)* 0.14 1.48 0.14 * ji) = 1i.58) 0.991, as compared to the assumed value of 0.99. 9. Comparisons with Design Values.a. Comparison of computed Uo with that given in the Happy Co. specifications. The computed Uo = 4.88 (based on outside area) = 669 (based on liner area). The overall coefficient specified is (Uo)specified = 98.4 (based on liner area) (Uo)computed 66.9 (Uo)specified 98.4 0 *Maxwell, J. B. Data Book On Hydrocarbons. New York: D. Van Nostrand Co., Inc., 1950, p.161..1 - - 130

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY: OF MICHIGAN b. Comparison of heat duties. Q 1,210,0o 007880 Qdesign 1,538,000 E. CALCULATION OF EXIT OIL TEMPERATURE, USING 651C 1, Air Velocity (see Section D-l). — Vface = 760 std ft/min Vmax = (760)(2.428) = 1845 std ft/min. 2, Outside Film Coefficient. —From Fig. 8, ho = 18.4 (based on outside area) (ho)L = (18.4)(13.7) = 252 (based on liner area) at a Vmax of 1845. 3. Metal Resistance.rm = 0.00062 (based on outside area). 4. Inside Film Coefficient.hi = 218 (A/[w) assume (l/tw) = 0.99 (to be checked later) hi = 216 (based on inside area). 5. The Inside Fouling is Assumed to be 0.002, Based on the Inside Area (See Section D-5). — 6. Overall Heat Transfer Coefficient — 1 1 15.28 U = 18.4 + 0.00062 + (15.28)(0.002) + 216 0.1564 Uo = 1 = 64 (based on outside area) 0.1564 131

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN (Uo)L = (6.4)(13.7) = 87.8 (based on liner area). 7. Calculation of Heat Duty. —Trial and error for the outlet temperature: assume toil (outlet) = 142.5~F Q = W Cp At = (90,659)(0.52)(173 - 142.5) = 1,4355000 Btu/hr. The corresponding temperature rise of the air is.tair.= _ __ 1,455,000 29.o0F Atair = (46,500) (60)(0.074) (0.24) = tair outlet = 100 + 29.0 = 129.0~F. The mean temperature difference is ATm (173 - 129.0) + (142.5 - 100) 452 ATM 43.2OF. 2 Check of heat duty: Q = Uo Ao ATm = (6.4)(5180)(43.2) = 1,430,000 Btu/hr, which checks closely the assumed value of 1,435,000 Btu/hr. 8. Check of Inside Coefficient (See Section D-8).(6.4)(15.28) toil = (6.4)(5.28) (43.2) = 19.5~F 216 toil at wall 173 + 142.5 = 138.2~F toil at wall = 2 ijw = 1.60 centipoise 0.14 1 4P 0.14 ( l) (. ) - 0o.989, as compared to the assumed value of 0.99. [ 132

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - 9. Comparison with Design Values.a. Comparison of computed Uo with that given in the Happy Co. specifications. computed Uo = 87~8 (based on liner area) specified Uo = 98.4 (based on liner area) (Uo)computed = 87.8 =. 89. (Uo)specified 98.4 b. Comparison of heat duties. Q _ 1,430,000 0.93Qdesign 1,538, 000 F. CALCULATION OF EXIT OIL TEMPERATURE, USING THE HAPPY CO. RATING CURVE 1. Air Velocity (See Section D-l).Vface = 760 std ft/min. 2. Outside Film Coefficients.-From Fig. 8, (ho)L = 220 (based on liner area) 220 ho = 116.05 (based on outside area). 1357 3. Metal Resistance.rm = 0.00062 (based on outside area). 4. Inside Film Coefficient.hi = 216 (based on inside area). 5. The Inside Fouling is 0.002, Based on the Inside Area.6. Overall Heat Transfer Coefficient.U 16.0o + 0.00062 + 0.03056 + 1568 = 0.1643 (based on outsided area) 153

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN U = 6.o8 (based on outside area) 0.1643 (Uo)L = (13.7)(6.08) = 83.5 (based on liner area). 7. Calculation of Heat Duty. —Trial and error for the outlet temperature: assume toil out = 143-5.F Q = (90,659)(0.52)(173 - 143.5) = 1,390,000 Btu/hr. The corresponding temperature rise of the air is Atai = 1,5390,000 = 28.10F r (46,500)(60o)(0074)(0.24): tair outlet = 100 + 28.1~F = 128.1~F. The mean temperature difference is AT = (173 - 128.1) + (143.5 - 100) = 44.2~F. 2 Check of heat duty: Q = (6.08)(5180)(44.2) = 1,395,000 Btu/hr, which checks closely with the assumed value of 1,390,000 Btu/hr. 8. Comparison with Design Values.a. Overall coefficients. (Uo)computed 83.58 0.848 or 85%o (Uo)specified 98.4 b. Heat duty. 1,390,000 =.905 = 90.5%. Qdesign 1,538,000

- ENGINEERING RESEARCH INSTITUTE. * UNIVERSITY OF MICHIGAN APPENDIX J CALCULATION OF DEW POINT AND BUBBLE POINT OF FEED TO THE EEBUTANIZER CONDENSER Inlet Vapor Composition (From Mr. W. F. Roberts of The Happy Co.). Component C3 (propane) i-C4 (isobutane) n-C4 (butane) i-C5 (isopentane) Mole % 1.40 21.80 75.20 1.60 DETERMINATION OF PHYSICAL PROPERTIES OF THE MIXTURE 1. Bubble Point (NGSMA, pp. 168-171). — Assume tb = 172 F: Component C3 i -C4 n-C4 i-C5 x 1.40 21.80 75.2 1.60 K 2.16 1.15.94.42 5.02 3.02 25.05 72.00 i.75 = 100.82 Therefore, the assumed tb = 172~F is correct. 2. Dew Point.Assume tD = 174~F: Component C3 i-C4 n-C4 i-C5 y 1.40 21.80 75.2 1.60 K 2.2 1.18.96.48 x=y/K 0.657 18.500 78.500 3-330 5 x = 100.767 the feed vapor Therefore, the assumed tD = 174~F is correct, and 135 is saturated.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN APPENDIX K CALCULATION OF TUBE WALL TEMPERATURE FOR PARTIALLY FILLED EVAPORATOR CONDENSER TUBE The relationship used to calculate the tube wall temperature is twall = tsv - Ati where twall tsv =the wall temperature, ~F, temperature of saturated vapors,.~F =240~F for this design, and =temperature drop across condensate and fouling film. )p, Ati, is related to the overall temperature difference by Ati The temperature drc Ati = (h + ri) ^i Aliner (Uo)L AToa where hi ri (Uo)L AToa = inside coefficient, based on liner area, = inside fouling film resistance, based on liner area, = overall coefficient, based on liner area, and = overall temperature difference = Tsv - Tair. Using: tair = 100~F (inlet air from specifications)* *The actual air temperature on the outside of the tube will be somewhat higher than the assumed value. The effect of increasing this temperature would be to increase the computed wall temperature of the tube. 136

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN (ho)L = 210 (based on liner area) from the Happy Co. rating curve ri = 0.0011 (based on liner area) from TEMA. The relationships given can be reduced to all = 40 -Lo(0.0011 hi + 1 twall = 24\0.00586 hi + lI) The following table can then be computed. TABLE I (APPENDIX K) COMPUTED WALL TEMPERATURES IN THE SUBCOOLING ZONE OF THE EVAPORATOR CONDENSER FOR VARIOUS INSIDE COEFFICIENTS hi twall, ~F 1000 194 1200 200 1500 202 2000 205 137

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN APPENDIX L CALCULATION OF FAN-BLADE EFFICIENCY Reference is made to Fig. 21 of page 58. For a 12~ blade angle, with a pressure drop of 0.4 in. of water, the anticipated flow rate is obtained as 275,000 CFM (p = 0.065 lb/cu ft). The corresponding brake horsepower delivered to the blades is obtained as 30 BHP. Computation of theoretical horsepower required: HP = (275,000't x 60 0mi5) 4 in. H20 x 62.4 l- b (5.05 x -7 Hr ) min hr 12 in./ft u ft H20 ft-lb = 17.35 HP Blade efficiency Theoretical HP required Actual HP required 17.35 30 *Maxwell, J. B., op. cit. p. 250. 138

UNIVERSITY OF MICHIGAN 1 I9IIII 101 lllllllI 3 9015 03524 9591