Boiler Surface Area: More is not Necessarily Better

The basic equation for heat transfer across a surface is:

Q = USDT where

Q = energy transferred
U = overall heat-transfer coefficient
S = surface area
DT = log mean temperature difference.

For the same DT, Q is affected by the quantity US, and not by S alone. In turn, the heat-transfer coefficient U in a boiler is a function of several variables, such as:

• Tube size-The lower the tube size, generally, the higher U is.
• Tube spacings-This affects gas velocity, which in turn affects U. Tube spacings also affect the heat-transfer-coefficient correction factor.
• Fin spacings-Fin density, height, thickness and fin conductivity affect heat transfer very significantly (a topic that will be explored at a later date). It can also be shown that more fin area increases the tube wall and fin tip temperature, forcing one to use higher-grade alloy steels for tubes and fins. Generally, the higher the fin surface area, the lower U is. Surface areas can vary by 100-300% for the same duty.
• Gas velocity-This affects U and gas pressure drop.

The example below shows how two boilers designed for the same duty with the same overall performance (back pressure and efficiency) differ in arrangement and surface area. The surface areas differ by about 9%. The difference would be much higher if a superheater were present. A radiant superheater requires less area than a convective design, and this distorts the picture even more. The table demonstrates that you can have same overall performance with different surface areas for each component (superheater, evaporator, furnace, etc). The example also shows that you can have comparable performance even when the the total surface area varies by 9% (for each boiler, total area = sum of furnace, evaporator, economizer areas, which total 13,158 vs. 14,336 sq. ft)

Also, the location of the convective superheater affects the surface area significantly. The company where I am employed, for example, prefers a convective-type superheater (located in a reasonably cool gas-temperature region), which increases the surface area of the superheater due to the lower log-mean temperature difference. However, the life of the superheater is generally longer compared to a radiant or semi-radiant design.

The conclusion is: Simply deciding on a particular boiler design or supplier simply because the design has more surface area can be wrong. Unless one can perform heat transfer calculations and check the surface area of each boiler, one should stay away from comparing surface areas. I stress this because with easy access to spreadsheets, even non-technical managers get into the act of comparing surface areas of vendors and arrive at wrong conclusions. One should look at the overall performance, operating costs and initial cost of boiler for comparison purposes.

Steam parameters:
Flow = 100,000 lb/h, 300-psig saturated steam
230°F feedwater
2 % blowdown
Natural gas fuel, 10% excess air
Boiler duty = 100.8 MM Btu/h
Efficiency = 84.3% Higher Heating Value (HHV)
Furnace back pressure = 7.0 in w.c.

By V. Ganapathy