
The number of transfer units (NTU) method is used to calculate the rate of heat transfer in heat exchangers (especially parallel flow, counter current, and cross-flow exchangers) when there is insufficient information to calculate the log mean temperature difference (LMTD). Alternatively, this method is useful for determining the expected heat exchanger effectiveness from the known geometry. In heat exchanger analysis, if the fluid inlet and outlet temperatures are specified or can be determined by simple energy balance, the LMTD method can be used; but when these temperatures are not available either the NTU or the effectiveness NTU method is used.
The effectiveness-NTU method is very useful for all the flow arrangements (besides parallel flow, cross flow, and counterflow ones) but the effectiveness of all other types must be obtained by a numerical solution of the partial differential equations and there is no analytical equation for LMTD or effectiveness.
To define the effectiveness of a heat exchanger we need to find the maximum possible heat transfer that can be hypothetically achieved in an ideal counter-flow heat exchanger of infinite length. Therefore one of the fluids would experience in that conditions the maximum possible temperature change, which is the difference of
Th,i- Tc,i
| m |
Ch=
| m |
hcp,h
Cc=
| m |
ccp,c
Here,
cp
the products
c
p= \partialh \partialT
| m |
cp
By conservation of energy, the total enthalpy change of both fluids must be the same (in absolute value) when they pass through the ideal heat exchanger. Therefore, the fluid with the smaller heat capacity rate will be the one that experience the maximum temperature change, whereas the other fluid would change temperature more slowly along the heat exchanger length. Therefore, the maximum possible heat transfer rate between the fluids is determined by the following expression:
| Q |
max =Cmin(Th,i-Tc,i)
where
Cmin=min(Ch,Cc)
Then, the effectiveness of the heat exchanger (
\epsilon
\epsilon =
| ||||
|
where the real heat transfer rate can be determined either from the cold fluid or the hot fluid (they must provide equivalent results):
| Q |
=Ch(Th,i-Th,o) =Cc(Tc,o-Tc,i)
Effectiveness is a dimensionless quantity between 0 and 1. If we know
\epsilon
| Q |
=\epsilonCmin(Th,i-Tc,i)
Then, having determined the actual heat transfer from the effectiveness and inlet temperatures, the outlet temperatures can be determined from the equation above.
For any heat exchanger it can be shown that the effectiveness of the heat exchanger is related to a non-dimensional term called the "number of transfer units" or NTU:
\epsilon=f(NTU,
| Cmin | |
| Cmax |
)
For a given geometry,
\epsilon
Cr
Cr =
| Cmin | |
| Cmax |
NTU
NTU =
| UA | |
| Cmin |
U
A
Cmin
| Q= |
Cmin(To-Ti)min=UA\DeltaTLM
From this energy balance, it is clear that NTU relates the temperature change of the flow with the minimum heat capacitance rate to the log mean temperature difference (
\DeltaTLM
For example, the effectiveness of a parallel flow heat exchanger is calculated with:
\epsilon =
| 1-\exp[-NTU(1+Cr)] | |
| 1+Cr |
Or the effectiveness of a counter-current flow heat exchanger is calculated with:
\epsilon =
| 1-\exp[-NTU(1-Cr)] | |
| 1-Cr\exp[-NTU(1-Cr)] |
For a balanced counter-current flow heat exchanger (balanced meaning
Cr =1
\epsilon =
| NTU | |
| 1+NTU |
A single-stream heat exchanger is a special case in which
Cr =0
Cmin=0
Cmax=infty
\epsilon =1-e-NTU
For a crossflow heat exchanger with both fluid unmixed, the effectiveness is:
\epsilon =1-\exp(-NTU)-\exp[-(1+Cr)NTU]
| infty | |
| \sum | |
| n=1 |
| n | |
| C | |
| r |
Pn(NTU)
where
Pn
Pn(x)=
| 1 | |
| (n+1)! |
| n | |
| \sum | |
| j=1 |
| n+1-j | |
| j! |
xn+j
If both fluids are mixed in the crossflow heat exchanger, then
\epsilon =\left[
| 1 | |
| 1-\exp(-NTU) |
+
| Cr | |
| 1-\exp(-NTU ⋅ Cr) |
-
| 1 | |
| NTU |
\right]-1
If one of the fluids in the crossflow heat exchanger is mixed and the other is unmixed, the result depends on which one has the minimum heat capacity rate. If
Cmin
\epsilon =1-\exp\left(-
| 1-\exp(-NTU ⋅ Cr) | |
| Cr |
\right)
whereas if
Cmin
\epsilon =
| 1 | |
| Cr |
(1-\exp\{-Cr[1-\exp(-NTU)]\})
All these formulas for crossflow heat exchangers are also valid for
Cr=1
Additional effectiveness-NTU analytical relationships have been derived for other flow arrangements, including shell-and-tube heat exchangers with multiple passes and different shell types, and plate heat exchangers.[2]
It is common in the field of mass transfer system design and modeling to draw analogies between heat transfer and mass transfer.[3] However, a mass transfer-analogous definition of the effectiveness-NTU method requires some additional terms. One common misconception is that gaseous mass transfer is driven by concentration gradients, however, in reality it is the partial pressure of the given gas that drive mass transfer. In the same way that the heat transfer definition includes the specific heat capacity of the fluid, a mass transfer-analogous specific mass capacity is required. This specific mass capacity should describe the change in concentration of the transferring gas relative to the partial pressure difference driving the mass transfer. This results in a definition for specific mass capacity as follows:
Here,cp-x=
d\omegax dPx
\omegax
Px
Here,cp-x=
Mx/Mother Pother
Mx
M other
Here,NTUx =
UmAm
• m cp-x
Um
Am
| m |
Cr
Cr =
• (m cp-x)min
• (m cp-x)max
One particularly useful application for the above described effectiveness-NTU framework is membrane-based air dehumidification.[4] In this case, the definition of specific mass capacity can be defined for humid air and is termed the "specific humidity capacity."
Here,cp-h=
Mwv/Mair = Pair
0.62198 = Pair
0.62198 Ptotal-Pwv,inlet
Mwv
Mair
Pair
Pwv,inlet
It is very common, especially in dehumidification applications, to define the mass transfer driving force as the concentration difference. When deriving effectiveness-NTU correlations for membrane-based gas separations, this is valid only if the total pressures are approximately equal on both sides of the membrane (e.g., an energy recovery ventilator for a building). This is sufficient since the partial pressure and concentration are proportional. However, if the total pressures are not approximately equal on both sides of the membrane, the low pressure side could have a higher "concentration" but a lower partial pressure of the given gas (e.g., water vapor in a dehumidification application) than the high pressure side, thus using the concentration as the driving is not physically accurate.