THEORY

The theory behind the operation of a double-pipe heat exchanger is covered in Incropera and
Dewitt (1996). Also in this same textbook is the derivation of how transient behavior is treated
with respect to heat transfer. As with any process the analysis of a heat exchanger begins with an
energy and material balance. Before doing a complete energy balance a few assumptions can be
made. The first assumption is that the energy lost to the surroundings from the cooling water or
from the U-bends in the inner pipe to the surroundings is negligible. We also assume negligible
potential or kinetic energy changes and constant physical properties such as specific heats and
density. These assumptions also simplify the basic heat-exchanger equations.
The determination of the overall heat-transfer coefficient is necessary in order to determine the
heat transferred from the inner pipe to the outer pipe. This coefficient takes into account all of
the conductive and convective resistances (k and h, respectively) between fluids separated by the
inner pipe, and also takes into account thermal resistances caused by fouling (rust, scaling, i.e.)
on both sides of the inner pipe. For a double-pipe heat exchanger the overall heat transfer
coefficient, U, can be expressed as

In a heat exchanger the log-mean temperature difference is the appropriate average

temperature difference to use in heat transfer calculations. The equation for the log-mean
temperature difference is
 Fluid properties such as density, viscosity and heat capacity are evaluated at the average
temperatures. The average is found by using the inlet and outlet values

Thermal conductivity, k, can be evaluated at the average of the average temperatures. In a

double-pipe heat exchanger the inner pipe is made of a conductive metal and is thin.

The problem can be further simplified if the equipment is assumed to be clean, which
means that no scaling exists. This is a poor simplification with the double-pipe heat exchanger in
the laboratory, because it is many years old. The fouling factors Rfo and Rfi can be looked up
from various sources, including Standards of the Tubular Exchange Manufacturers Association,
or lumped together and determined experimentally.

The only part of the overall heat-transfer coefficient that needs to be determined is the
convective heat-transfer coefficients. Correlations are used to relate the Reynolds number to the
heat-transfer coefficient. The Reynolds number is a dimensionless ratio of the inertial and
viscous forces in flow.

In the inner pipe if the Reynolds is less than 2000 this is considered to be laminar flow
and the Nusselt number is equal to 4.36. If the Reynolds number is greater than 10,000, the
Nusselt number is given by
 This gives a Nusselt number that can then be use to find hi

The convective heat transfer coefficient in the annulus is more difficult to determine. The
hydraulic diameter is used to find the Reynolds number. The hydraulic diameter is defined as the
cross-sectional area perpendicular to flow divided by the wetted perimeter. With the Reynolds
number calculated the same correlations apply and with these ho can be determined.

Once all the separate heat-transfer coefficients are calculated an overall heat transfer
coefficient is calculated. Now everything that was necessary for an energy balance is available.

With the previous assumptions made earlier the dynamic equations would be

With the transient data taken experimentally an overall heat-transfer coefficient can be
determined at each time step. This can be solved numerically.