Changing Tube-Side Passes

Once my client in South Africa changed a water cooler from two-pass tubes to four-pass tubes (see Fig. 32.1), meaning that the water traveled through half the number of tubes per pass. The water traveled twice as far, as it now went through the tube bundle four times. 

Let s say your sea-water pumps develop a 40-psig discharge pressure. The sea-water return line operates at 5 psig. The pressure drop 

When my client in South Africa changed from two-pass to four-pass, the water flow decreased from 1,000,000 Ib/hr to 350,000 Ib/hr. The method of calculation to derive the lower flow is somewhat complex. The calculations take into account that the water has to flow twice as far through the tube bundle, and that there are only half the number of tubes available per pass, but that the water pressure drop must remain constant. 

Due to the lower flow, the temperature increase of the sea water also went up  

At 150°F, the cooling water precipitated hardness deposits. The hardness deposits restricted water flow. The restricted water flow increased the water outlet temperature. The increased water temperature reduced the water flow. The reduced water flow... but perhaps I have made my point. We changed back to the two-pass configuration. Kindly compare this story to Chap. 26, Shell-and-Tube Heat Exchangers where tube passes were increased. 

---

Calculate the pressure drop for the water flowing through the air-cooled heat exchanger designed in Example 7.37 if the number of tube-side passes is 10. The density of the water is 60 lb/ft3 (961.1 kg/ m3), and the viscosity is 0.74 lb/(ft)(h) (0.31 cP). Assume that the velocity in the nozzles is 10 ft/s (3.05 m/s) and that the viscosity change with temperature is negligible. 

If the tube-side pressure drop exceeds a critical allowable value for the process system, then recheck by either lowering the flow rate and changing the temperature levels or reassume a unit with fewer passes on tube side or more tubes per pass. The unit must then be rechecked for the effect of changes on heat transfer performance. 

The velocity on the tube side can be modified by changing the single-pass design to a multiple-pass configuration. In this case Ft = 1 in Equation (b). From formulas in McCabe, Ft depends on t2 (or At2), hence the necessary conditions derived previously would have to be changed. The fluids could be switched (shell vs. tube side) if constraints are violated, but there may well be practical limitations such as one fluid being quite dirty or corrosive so that the fluid must flow in the tube side (to facilitate cleaning or to reduce alloy costs). 

An even number of tube passes is desireable because then input and output of the tube-side fluid occur on the same end of the exchanger. This allows the other end of the exchanger to "float" in response to thermal expansion of the metal, which can otherwise be a serious problem when the shell changes temperature during startup or shutdown. 

For this formulation to apply to shell-and-tube exchangers, it is assumed that each tube pass has the same number of tubes and that the tube-side coefficient in each pass is the same. The analysis also requires that there be sufQcient baffles that the shell-side flow can be treated as longitudinal to the tubes, rather than as a series of cross-flow sections. Three or four baffles may be sufficient to meet this criterion if the shell-side temperature change is less than the minimum temperature difference between shell-side and mbe-side fluids, and eight or more baffles are almost always sufficient for the assumption to be satisfied. 

If none of these three suggestions help, increase tube-side mass velocity. The overall tube heat-transfer coefficient is largely a function of the process fluid mass velocity. On one unit, changing from four parallel passes to two parallel passes increased the rated furnace capacity by 20%. Maintaining a high mass velocity flb/ftVsec) is the best way to keep tubes cool. Rating furnace capacity on the basis of heat flux (BTU/hr/ft ) without consideration of tube mass velocity is a serious mistake. 

---