1|Page

8 Design Considerations Double-pipe Exchangers

These are the simplest of exchangers. They can be easily assembled from standard parts in any workshop. Their cost per ft2 is relatively small. Also, the correction factor F = 1. When we speak of the diameter of a pipe, we mean the nominal diameter. C-steel pipes are used in most services. The sizes of steel pipes conform to the Iron Pipe Size (IPS) standards. For pipe sizes up to and including 12, the nominal diameter corresponds to neither the inner nor the outer. Small sizes result in large values of Re and consequently of h, If the pressure drop proves excessive then larger sizes should be considered during design. For a given nominal diameter, the OD is fixed while the wall thickness and hence the ID vary according to Schedule number. Schedule numbers 40 (standard) and 80 (extra heavy, used with operating pressures > 50 atm.) are generally used with pipe sizes < 12 and therefore for double-pipe pairs. The wall thickness is seen to increase with the schedule number. Effective lengths of 12, 15, or 20 are commonly used: Beyond 20, the inner pipe tends to sag and touch the outer pipe. Double pipes may be joined together in the form of hair pins The disadvantages of double pipes are that leakages are very common, and they require a lot of time for dismantling and cleaning. Also the amount of surface per double pipe is small; when a large number of hair-pins are used the space requirement is considerably large. The outer surface of the inner pipe may be increased by the use of longitudinal fins and they are used when (ho<1/2 hi), they are then more economical. Double pipes should be considered during a design. They are used when the required heat transfer surface does not exceed 100-200 ft2 If several double pipes are used the weight of metal required for the outer pipes becomes so large that the shell-tube construction is then more economical.

Shell-tube Exchangers
The shell-tube exchanger consists essentially of a number of tubes surrounded by a shell with or without baffles. Extensive information pertaining to this exchanger type is furnished by the TEMA standards, Turbular Exchanger Manufacturers Association The 2-4 exchanger could consist of a single unit or two 1-2 exchangers with their shells and tubes connected in series

Heat Exchanger Tubes

These should not be confused with steel pipes or other types of pipes which are extruded according to iron pipe sizes. When we speak of the diameter of a tube we always mean the outside diameter. Tubes are obtained in a number of wall thicknesses defined by the Birmingham Wire Gage (BWG) The wall thickness is seen to decrease as the BWG increases. The tube-side heat transfer coefficient is favoured by small tube inside diameters; however it should be borne in mind that the corresponding pressure drop increases as well.

2|Page

In Egypt, 16 tubes are the most commonly used. 8 tubes sharply increase the 1st cost per ft2. Long tubes are difficult to handle during cleaning and maintenance of the exchanger especially at high elevations; 4 or8 tubes are used in this respect. Tubes may be arranged on triangular or square pitch layouts. With triangular pitch, the tubes can be placed more compactly) within a given shell size, and the shell side De is smaller; the corresponding heat transfer surface and shell-side coefficient are larger than for square pitch. On the other hand, square pitch offers the following advantages: The outer surface of the tubes is more accessible for cleaning with tools, and the shell-side pressure drop is comparatively smaller than for triangular pitch. With highly fouling or viscous shell-side fluids square pitch is indicated, otherwise triangular pitch is preferable. When using more than 8 tube passes in large shells (31 -39), the tube count for 8 passes should be reduced by 5% for 12 passes and 10% for 16 passes. For smaller shells, it is advisable to avoid the use of 12 and 16 passes.

Shells
They are fabricated from steel pipe with nominal IPS diameters from 8-24. The standard wall thickness in this case 3/8 which is satisfactory for shell-side operating pressures up to 300 psi; larger wall thicknesses may be obtained for greater pressures and the shells are then more costly. Shells above 24 in diameter are fabricated by rolling steel plate; the thickness in this case is 7/16.

Baffles
Transverse baffles are used to increase the turbulence and hence the heat transfer coefficient within the shell. Segmental baffles, notably 25% cut ones, are the most commonly used. They consist of circular discs of sheet steel with one side cut away, and perforated to receive the tubes. The baffles may be arranged for up-and-down flow or they may be rotated 90 to provide side-to-side flow the latter arrangement is used when a mixture of liquid md gas, for example condensing vapors, is passed through the shell; the liquid would be trapped behind the baffle if the flow were up-and-down. The closer the baffle spacing, the smaller the available flow area, and consequently the higher the shell-side Re and the corresponding values of ho and Po. Baffle spacing larger than the maximum is practically ineffective, while spacing closer than the minimum leads to excessive pressure drop. Excessive clearances lead to leakages. To avoid vibration of the baffles and scoring of the tubes, the baffles should be at least 1/8 thick or twice as thick as the wall whichever is thicker. Longitudinal baffles, if used, may be welded to the shell or may be removable with seals. Dam baffles are used in Horizontal condenser sub-coolers to provide sub-cooling surface

3|Page

Routing of the fluids

1. High pressure fluids, more fouling or more corrosive fluids are routed in the tubes. 2. More viscous fluids or low flow rate fluids (low Re) are routed in the shell.. To avoid excessive shell side pressure drop, baffles might have to be omitted in the case of highly viscous fluids. 3. Larger flow rate stream is routed in the larger cross sectional area, so as to bring both fluid mass velocities close to one another. 4. Cooling water or heating steam should be routed in the tubes unless the fluid being cooled or heated is under comparatively high pressure or if it is highly fouling.

Cleaning of heat Transfer Surfaces

Methods of cleaning may vary according to the type of deposits. 1- Hard Deposits a) Rust b) Petroleum coke c) Scale (mineral salts(Ca(OH)2, CaSO4, Na2SO4, Na2CO3 or SiO2)) Hard Deposits are cleaned by 4 ways i. ii. Sand blasting Cleaning tools: Rotary cleaners +air current from inside tubes Scraping bars, stiff wire brushers, water under high pressure for outside tubes.

Cleaning of the inside of the tubes is simpler than cleaning of their outside surfaces, since the use of rotary cleaners requires considerably less efforts than other tools. Also the tube bundle needs not to be removed from the shell. This is why it is preferable to route the high fouling fluid within the tubes of the shell & tube heat exchanger. iii. Circulating Chemicals Phosphates such as sodium hexa meta phosphate (calgon/calgonite) Dilute HCl + inhibitor (to prevent corrosion) Sodium or Ammonium fluoride (with scales high in SiO2)

iv. Boiling water to remove Na2SO4and Na2CO3 The exit temperature of cooling water should not exceed about 50oc to prevent the excessive decomposition of mineral scale. Thus Ca(OH)2 is more soluble in cold water than in hot CaSO4 is sparingly soluble in cold water; its solubility increases with Temperature up to 40-50oc after which it falls forming a very hard scale.

4|Page

2- Soft Deposits such as oil films i. ii. Sand blasting Boiling Alkaline solution

3- Loose Deposits such as leaves, algea, silt, soft lime i. ii. iii. iv. v. Sand blasting Blowing air or steam Flushing with hot water Soft lime may be removed from inside of tubes by driving a rubber plug through the tube with the help of water under pressure. It is recommended that tube side water velocities should exceed 3-4 ft/s so that loose deposits are carried away.

Shell and Tube Passes

The simplest of shell-tube exchangers is the It offers the advantage of counter flow, i.e., F =1. For a given shell size the tube-side mass velocity is low compared to multi-pass exchangers resulting in lower heat transfer coefficients and pressure drops. Multi-pass exchangers could be referred to as p-2p (p shell passes irrespective of baffle presence-2p or more even numbers of tube passes) exchangers. Usually the maximum number of tube passes is 8 although for large shells (31-39) up to 16 passes may be used. Thus, for example, a l-2 exchanger consists of 1 shell pass and 2, 4, 6, or 8 tube passes. The values of F for a 1-2 and 1-8 exchanger are less than 2% apart in the extreme case and generally considerably less. It is therefore customary to refer to any exchanger having 1 shell pass and two or more even numbered tube passes in parallel flow-counter flow as a 1-2 exchanger. Similarly a 2-4 exchanger consists of 2 shell passes and 4 or more even numbered tube passes, and so on. Of the shell, tube exchangers, the multi-pass exchangers especially the 1-2 exchanger are the most commonly used. For a given shell size, tube diameter and configuration, the number of tubes per tube pass is smaller than for a 1-1 exchanger; the mass velocity is thus larger, and consequently the tube-side heat transfer coefficient and pressure drop are larger. Although the overall coefficient is favored due to the increase of hi, the temperature difference correction factor is less than unity. The floating head, which provides the most satisfactory solution to exchanger differential expansion problems, is commonly used with multi-pass exchangers. In the case of 1-1 exchangers the floating head is difficult to realize, and the exchanger is then subject to leakage. Sometimes it is impossible to meet the allowable pressure drop requirement. This may occur when the mass flow rate of one fluid is extremely large and the corresponding allowable P value is very small. Such situations are especially encountered in the case of gases and vapors; typical values of Pall. Because of the high mass velocity involved, the calculated value of the combined fouling factor is then usually much larger than the allowable value. The fluid pressure drop rather than the fouling factor becomes the controlling factor and one can afford to reduce the former at the expense of the latter. The idea is to reduce the mass velocity and the friction length.

5|Page

The pressure drop is affected more strongly by mass velocity than is the heat transfer coefficient. The following solutions to the problem of pressure drop may then be adopted: Tube-side: The use of a 1-1 exchanger rather than a multi-pass exchanger is indicated. Shell-side: The flow could be split into two parallel streams inside the shell of the exchanger

Process & Exhaust Steam

Some plants obtain power from non-condensing steam turbines or engines. In such plants there may be an abundance of exhaust steam at 5-25 psig (saturation temperature 228-267 F) which could be used for heating purposes. If a fluid is to be heated to a temperature close to or above the temperature of the exhaust steam, process steam at 100-200 psi (saturation temperature = 328 382F) developed at the power house especially for process purposes could be used. Process steam cost 4-8 times as much as exhaust steam. When process steam is used, all the heating can be done in a single shell using process steam. A better alternative is to distribute the heat load among two exchangers with the cold fluid sides connected in series, the 1st exchanger using as much exhaust steam as possible and the second using as little process steam as possible so as to minimize costs.