Page 111 - Applied Process Design For Chemical And Petrochemical Plants Volume III

66131_Ludwig_CH10C 5/30/2001 4:21 PM Page 78

78 Applied Process Design for Chemical and Petrochemical Plants

Edmister and Marchello 165 present a tube wall tempera-
h i t i h o t o
t w ture equation:
h i h o
(10-30)
where t 2 D i h i t i D o h o t 4 , °F (10-34)
t i bulk temperature of fluid inside tube D i h i D o h o
t o bulk temperature of fluid outside tube
where
h i film coefficient for fluid inside tube
t 1 fluid No. 1 mean fluid temperature, °F
h o film coefficient of fluid outside tube
t 4 fluid No. 2 mean fluid temperature, °F
t 2 fluid No. 1 fluid film tube wall temperature, °F
To determine a reasonably good value for t w , either t w must t 3 fluid No. 3 fluid film tube wall temperature, °F
be estimated and used to calculate h i and h o , or h i and h o must D o tube outside diameter, in.
be assumed and a t w calculated. In either case, if the calcu- D i tube inside diameter, in.
lated values do not check reasonably close to the assumed val- h i inside tube film (surface) coefficient,
2
ues, the new calculated results should be used to recalculate Btu/(hr) (ft ) (°F)
better values. Film temperatures are generally taken as the h o outside tube film (surface) coefficient,
2
arithmetic average of the tube wall, t w , and the bulk tempera- Btu/(hr) (ft ) (°F)
ture, t i or t o . This approach neglects the effect of tube wall
This relationship can be used for estimating surface tem-
fouling (i.e., it is for clean tube conditions). Corrections can
perature and for back-checking estimating assumptions. For
be made to account for the fouling if considered necessary.
many situations involving liquids and their tube walls, the
Usually this fouling is accounted for in the overall U.
temperature difference (t 2 t 3 ) across the wall is small and
The temperature for calculating film properties is as follows.
equals t 2 t 3 for practical purposes.
For streamline flow:
Fouling of Tube Surface
t f t av 1>4 1t w t av 2 (10-31)
For turbulent flow: Most process applications involve fluids that form some
type of adhering film or scale onto the surfaces of the inside
t f t av 1>2 1t w t av 2 (10-32) and outside of the tube wall separating the two systems (Fig-
ure 10-28). These deposits may vary in nature (brittle,
Ganapathy 161 presents a shortcut technique for estimating gummy), texture, thickness, thermal conductivity, ease of
heat exchanger tube wall temperature, which so often is removal, etc. Although no deposits are on a clean tube or
needed in establishing the fluid film temperature at the bundle, the design practice is to attempt to compensate for
tube wall: the reduction in heat transfer through these deposits by
considering them as resistances to the heat flow. These resis-
t w 1h i t i h o t o 2 > 1h i h o 2 (10-33) tances or fouling factors have not been accurately deter-
mined for very many fluids and metal combinations, yet
where general practice is to “throw in a fouling factor.” This can be
h i inside tube coefficient, Btu/(hr) (ft ) (°F) disastrous to an otherwise good technical evaluation of the
2
h o outside tube film coefficient, Btu/(hr) (ft ) (°F) expected performance of a unit. Actually, considerable
2
t i temperature of inside fluid entering tubes, °F attention must be given to such values as the temperature
t o temperature of outside fluid on tubes °F range, which affects the deposit, the metal surface (steel,
copper, nickel, etc.) as it affects the adherence of the
For example, a hot flue gas flows outside a tube and shell deposit, and the fluid velocity as it flows over the deposit or
exchanger at 900°F (t o ) while a hot liquid is flowing into the else moves the material at such a velocity as to reduce the
tubes at 325°F (t i ). The film coefficients have been estimated scaling or fouling.
2
to be h i 225°F and h o 16 Btu/(hr) (ft ) (°F). Estimate The percentage effect of the fouling factor on the effec-
the tube wall temperature using h i as h io corrected to the tive overall heat transfer coefficient is considerably more on
outside surface for the inside coefficient: units with the normally high value of a clean unfouled coef-
ficient than for one of low value. For example, a unit with a
t w 312252 13252 1162 190024 > 1225 162 363°F clean overall coefficient of 400 when corrected for 0.003
total fouling ends up with an effective coefficient of 180, but
This calculation neglects the temperature drop across the a unit with a clean coefficient of 60, when corrected for a
metal tube wall and considers the entire tube to be at the 0.003 fouling allowance, shows an effective coefficient of
70
temperature of the outside surface of the wall, t w . Kern for 50.5 (see Figure 10-39).
the same equation suggests using the caloric temperature Fouling factors as suggested by TEMA 107 are shown in
for t i and t o . Table 10-12. These values are predominantly for petroleum