Document rp90dx9j2Ybe87m363mneK9er

Reprinted from CHEMICAL ENGINEERING PROGRESS, May 1963 High temperature heating media Therminol FR-2 heat transfer systems Engineering design data are given for the design of high temperature systems using chlorinated biphenyls. Therminol fr-2 is a mixture of chlorinated biphenyls used as a heat transfer fluid to a maximum bulk temperature of 600F because of its excellent thermal stability, fire-resist ance, and high boiling point. The fluid was previously designated Aroclor 1248. Physical properties of this heat transfer fluid are given in Table 1. Thermal stability When Therminol FR-2 is heated ibove 600F, the fluid forms polymers af chlorinated biphenyl with the liber ation of gaseous HC1. The higher molecular weight polymers are mod erately soluble in the fluid, and the only measure of thermal degradation is the increased viscosity. Figure 1 shows the results of static thermal stability tests carried ont under isothermal conditions. W. J. Davis and P. G. Benigntjs Monsanto Chemical Co. St. Louis, Mo. The polymer formation was found to be first order following Equation 1 until the viscosity increased from 20 centipoises to 30 centipoises measured at 130F. bg(VJVt) = 0.434 K,f (1) Using Equation 1 together with the experimental rate constants, the inter mediate viscosity curves for other tem peratures were plotted on Figure 1. When the calculated values deviated from the experimental curves, inter mediate values determined from special cross-plots were used to smooth the curves. CHEMICAL ENGINEERING PROGRESS, (Vol. SI. No. S) DSW 321728 May 1963 39 STLCOPCB4073379 The equivalent cold plane surface aAcp is the product of a plane A^ re placing the heater tube bank and an effectiveness factor which is the ratio of radiation reception by the actual tube surface to die reception by the continuous plane. A plot of a based on data by Hottel is presented in Mc Adams^). Maximum heat flux Equation 2 together with the over all heat balance, Equation 4, can be solved for the maximum heat flux QJaAcf by assuming values for Ta until Equation 4 is satisfied. Assume 2% of the heat input as losses. @(Input) = Q. (Absorbed) -f- @0(Exhaust Gases) -f- QL Figure 1. Effect of temperature and time on the viscosity of Therminol FR-2. When fluid bulk or film tempera tures exceed 630F, the fluid viscosity increases rapidly with temperature. In a properly designed commercial heater operating at a bulk temperature of 600F, with film temperatures below 630-640F, the over-all rate of poly mer formation is extremely low. This conclusion is substantiated from field results of several operating heaters. Heater study A field study of several Therminol beaters was made to establish engineer ing design data such as allowable heat fluxes, maximum film temperatures, fluid stability rates, etc. The test heat ers included ones where tube failures occurred shortly after startup and heat ers which had been in successful oper ation for a number of years. The Lobo and Evans (I) method seemed most applicable to evaluate heater performance since it had al ready been used to test many heaters. The method was based on the Stephan-BoItzmann equation and previous ly published data by Hottel and co-workers (2, 3). More rigorous methods published prior to 1958 ap peared too rigorous for easy applica tion and the rough data available did not justify a more rigorous approach. A method was needed to distinguish between a heater which would operate satisfactorily and one which would be a problem, even though the method was empirical. Equation 2 is the Stephan-Boltzmann equation used by Lobo and Evans with an added convection term. QA = 0.Yt3 4,l(Ta/10Q)*- <f> is dependent upon the flame emissivity, eF, and the ratio of the effective refractory surface AR (total refractory surface At less aAcp) and the'equivalent cold plane surface aAcp. Lobo and Evans presented a <f> factor plot in their article. Flame emissivity has been de fined by Hottel (4) as: fr -- 1*q -- ceg(Tg/Tg*]/ [1 - (Tb/T0)*] (3) Values of the combustion gas emissivities and absorptivities can be obtained from radiation charts (4) for C02 and HzO. ,, -i .-4 (Heat Losses) (4) In applying the previous equations to evaluate the field test heaters, a trial value for Q(input) was assumed. If the calculated absorption rate did not agree with the actual rate within 1% after Equations 1 through 4 were satisfied, a new value for Q( input) was used until the calculated and the actual duties were equal. This pro cedure yielded values for T0 and QJ tzAcv maximum. Heater efficiency esti mated by this procedure is incorrect due to the basic assumption that the radiating gas temperature Ta is uni form throughout the heater. The final value of Ta must be adjusted by sub tracting 200 to 250F to Ta for esti mating the heat content of the com bustion gases leaving the heater (5). (T,g/100)4]aACJ) + hcA(T0-Ts) (2) A Therminol FR-2 high temperature heater installation. <C?>4 40 May 1963 CHEMICAL ENGINEERING PROGRESS. (Vol. 5?. No. 5) STLCOPCB4073380 CEP feature Maximum film temperatures were study were analyzed with both single estimated by calculating the outside zoned and multi-zoned calculations to fluid heat transfer coefficient using determine the film temperatures. The Equations 5 and 6 and solving Equa- majority of the film temperatures cal ton 7 for the temperature of die film, culated using a lower zone L/D of 2 Tf. agreed with the fluid life data and with h(Di/Jt=0.027 " comparable results obtained in box (Cr/kWrS'pjK" hfg = ht{Af/Ag) (5) (6) Qa/ol = (Q/A)ma, = hi0AT = Wr,-rt) (7) heaters. When film temperatures ap proached or exceeded 700F, tube burn-out or severe fluid damage oc curred. Excellent fluid life was ex perienced in heaters where the film temperatures were .below 640F. Table 2 is a summary of the data obtained on several of the test heaters. The tubes in units 1 through 4 burned out shortly after startup. The calcu The design of a liquid heating sys tem can be broken down into two major parts. These are: 1. heater de sign, and 2. system design. lated film temperatures in each case were approximately 700F. Inspection Heater design of Figure 1 shows that the fluid de There have been two types of composition rate at 700F temperature heaters used for Therminol FR-2 serv is very rapid. High polymeric tars ice. These are the floor-fired cylin rapidly form which settle out and drical heater and the box heater. Each cause blockage of orifices, control is of the forced-circulation type; valves, headers, etc., and eventual tube natural circulation heaters should not failures. be used because of the non-boiling Units 5 through 17 represent heat characteristics of the fluid. Forced ers which have been in successful circulation provides positive fluid cir- operation for the number of years des -.culation for controlling the film tem ignated and where the fluid is still the perature of the fluid near the coil original charge. In each case the esti surface. mated film temperatures were 640F A single coil heater is preferred to or less which correlates very well with the multi-pass heater from the stand the thermal stability data on Figure 1. point of positive circulation. How Cylindrical heaters ever, by proper design of the multi tube units to assure uniform distri Lobo and Evans' method was based bution of liquid flow, excellent per on the well-stirred furnace chamber formance can be obtained. In many model. Kern (6) recommended that in cases pressure drop across the unit applying the method to cylindrical must be low and multipass tube ar heaters that the unit be zoned with rangements even in cylindrical heaters a maximum zone length of 1.5 coil must be used. diameters. The use of a convection section in It can be shown that calculated flux creases the heater efficiency from 60 rates are much higher in the lower 65% to 75-78%, depending upon the half of a cylindrical heater if the unit design. Although the initial cost of is zoned rather than treating it as an the heater is significantly higher with isothermal single zone. For this reason the added convection section, usually each of the cylindrical heaters in the the fuel savings yields a short write TAMC I A- ,, 1 1 .? t'T l4=L i r USER I | STSTEM PUKE l J| ^.Jl C\ U LXE l IsECOHOUTY LOOP 'Q Jfl C00UMC PUMP * COOLER CT :; -SJ rd < MUM LOOP i i LOOP . 1 CONNECTIM Wfi. J 4-i **5--------------- ---------------------------------- --------------------------------- X------ pump @ Figure 2. Typical system for heating and cooling cycles with Therminol FR-2. High temperature heating media off period. In many cases where a convection section is being used, it is advantageous to introduce the cold fluid into the radiant coil rather than to use the conventional countercurrent flow pattern. The AT across the coil is usually 60F or less and with a maximum fluid limit of 600F, little efficiency will be lost due to the re duced temperature approach. How ever, the effect is significant in the radiant section, because lower bulk temperatures permit higher heat fluxes to maintain die maximum film tem perature. Where excessive decomposi tion is occurring in existing heaters, reversing the fluid in this manner would be most beneficial with very little cost involved. The Lobo-Evans' method (I) al ready discussed, can be used for rat ing heater designs. The maximum allowable film temperature recom mended is 640F based on the max imum flux rate (Q/aAep) determined by the method. Cylindrical heaters should be treated with a 2-zone cal culation (6) with the length of the lower zone equal to 2D. The zone calculation results in slightly higher coil areas to allow for uneven tem perature distribution in the heater and is not representative of actual thermal conditions. Box heaters can be treated with a single-zone calculation. Burner selection The type of burner to be used should be carefully selected by the designer. It has been found that the low pressure combination oil-gas forced draft unit is the most satisfac tory unit both from design and opera tional considerations. The initial cost is many more times that of the high pressure, natural induced type, but the advantages of better temperature control due to its higher turn-down ratio can readily be seen, especially when the heat load fluctuates widely. Most forced drafts units operate at turn-down ratios of 9:1 compared to 4:1 for the induced type. Since each application differs, it should be re membered that the burner is an im portant component of the system and should be carefully selected for the particular application. CHEMICAL ENGINEERING PROGRESS, (Vol. 57, No. 5) 0S\N 321730 May 1963 41 STLCOPCB4073381 System design flow rate through the heater. It is Figure 2 illustrates a typical heat ing system. Major items are the heater, circulating pump, vent tank, primary and secondary loops, and control in strumentation. The circulating pumps should be high temperature continuous duty steel centrifugal-type pumps. The pump should be installed near the heater at floor level for easy main or cooling water in the same jacket tenance accessibility. Either indoor or or coil. . outside installations can be used. The main circulating loop supplies The vent tank can be sized either hot fluid to the users and returns the to allow for system expansion or to cold fluid to the heater. It is not contain die entire charge of fluid. It recommended even in small installa can be fabricated from mild steel and tions that the users be placed in series should be insulated to prevent con with the loop. Always install the user densation of water vapors and HC1 in parallel or provide a by-pass around above the liquid level. A 1-in. surge the user. line should be installed connecting Direct fired heaters are usuallv pro the high temperature--high pressure vided with instrumentation to obtain: side of the main circulating loop to 1. adequate operational control and 2. the vent tank as shown in Figure 2. safety controls which will satisfy Fac Hot fluid should be recirculated to tory Insurance Association (FIA) ap the vent tank (above liquid level) con proval for direct-fired heaters. tinuously or intermittently to remove Figure 2 shows the fluid tempera traces of water and HC1. Caution ture-controller output regulating the should be exercised never to use fuel supply and the flow by-pass con Therminol intermittently with steam troller which maintains the design ................................................................... .......................................... . Table 1. Physical properties of Therminol FR-2. strongly recommended that the by pass control valve be installed close to the furnace to minimize time lags and excessive pressure drop in the loop piping. The recirculation rate through the heater does not necessar ily have to be the same as through the piping system; in most cases, it is larger. These two controllers are the most important portion of the in strument package to assure troublefree heater performance. A booklet on good practices has been published by the FIA (7). Ap proval by FIA is usually obtained during the design stage of any directfired heater installation. ACKNOWLEDGMENTS The authors wish to extend their personal gratitude to Monsanto Chem ical Co. for their permission to pre sent these data. They particularly wish to thank A. M. Ellenburg and H. D. Bamstarff for their much ap preciated data on thermal stability. Notation A--area, sq. ft C--fluid specific heat Btu/Ib.-0F. D--diameter, ft. G--mass flow rate, Ib./hr.-sq-ft. Temp., F 100 150 200 250 200 250 400 450 600 550 600 Sp. ht., Btu/lb-'F 0.281 0.289 0.297 0.305 0.313 0.321 0.329 0.337 0.333 0.341 0.349 Thermal Conductivity, Btu/hr.-F-ft. 0.0585 0.0580 0.0579 0.0575 0.0570 0.0566 0.0660 0.0555 0.0549 0.0542 0.0536 Density, lb./gal. 12.00 11.75 11.63 11.29 11.06 10.83 10.60 10.37 10.14 9.90 9.67 Viscosity, cp. 62.4 11.50 4.95 2.87 1.87 1.32 0.99 0.78 0.63 0.53 0.45 hI--rate constant. he--convective heat transfer coefficient Btu/hr.-F-sq.-ft fij--tube-side fluid heat transfer coeffi cient, Btu/hr.-F-sq.-ft. hfg--tube-side fluid heat transfer coeffi cient based On out-side tube area, Btu/hr.-F-sq.-ft. A:--fluid thermal conductivity, Btu/hr.F-ft. f--time, hr. T0--temperature of the combustion gases, R. Tg~tube wall temperature, R. Table 2. Therminol FR-2 field heater study. V#--initial fluid viscosity, cp. Vt--fluid viscosity at time (f), cp. Unit Capacity, Btu/hr. 1 6,000,000 2 20,000,000 8 5,000,000 4 3,000,000 6 6,000,000 6 2,000,000 7 1,000,000 8 8,000,000 9 6,000,000 10 10,000,000 11 5,000,000 12 3,000,000 13 1,500,000 14 3,000,000 15 -------8,000,000 16 6,000,000 17 9,000,000 B. O. Burn out. Bulk temp. 600 600 430 600 600 600 530 - 650 585 475 550 650 600 570 - 635 600 690 Flux BTU/HR.-8Q. FTi Max. Avg. 19,200 9,950 22.400 10,000 21.400 13,700 18,350 7,550 19,200 9,950 48,800 20,000 7,400 3,600 12,900 6,700 17,850 9,300 23,300 13,200 21.400 11,600 18,350 7,550 10,000 6,400 19,150 9,950 27.400 17,800 19,600 12,500 Max. FILM TEMP. P --Fluid VELOCITY FT./SEC. Fluid LIFE TO DATE, YE. 696 3.9 B.O. 705 3.9 B.O. 694 1.6 B.O. 726 5.0 B.O. 640 7.8 4 675 8.6 7 593 2.5 11 627 -- 8.8 8 631 10.3 4 573 3.2 4 631 4.0 7 640 6.7 8 635 5.8 6 640 8.0 5 595 12----- ------ 2----- 656 9.7 2 646 6.0 1 Greek letters -. to--gas emissivity. p--fluid viscosity, Ib./hr.-ft ^-over-all furnace emissivity. LITERATURE CITED I. Lobo. W. E.* and J. F. Evans, Tram. Am. /mt. Ckem. Engr$. 85, p. 743-78 (1959). 2. Hottel, H. C., Trans. Xm, Inst. Cham. Engrs^ IP, p. 173 (1927). 9. Hottel, H. C,, Trans, Am. Sac. Meeh. Enors., 58. (14), p. 265 (1931). 4. McAdunm, W. H., "Heat Transmission." McGraw-Hill Book Co., 2nd Ed. p. 70 <1942). 5. Hottel, H. O, /. inst. fW, p. 220-234 (1961). 6. Kern, D. Q., "Process Heat Transfer/' Mc Graw-Hill Book Co.. 1st Ed. (1950). 7. "Recommended good practice for beat transfer mediums in closed systems," Fac tory Insurance Association. Chicago, lit (June, 1964). 42 May 1963 DSW 321731 CHEMICAL ENGINEERING PROGRESS, (Vol. St. No. S) STLCOPCB4073382