Abstract
Cross-flow staggered tube bank heat exchangers are widely preferred in heat recovery and dry-type cooling towers due to their ease of cleaning and high volumetric heat transfer. This study performs an economic optimization for such a heat exchanger located in Turkey's 2nd climate zone (Gaziantep). The analysis considers initial investment costs (pumps, fans, and the heat exchanger itself) and operating costs (electricity consumption for pumps and fans) calculated at present value. The results, derived across various parameters, identify optimal design configurations to minimize total lifecycle costs.
Introduction
Cross-flow staggered tube bank heat exchangers are industry workhorses, frequently used in closed-type cooling towers and heat recovery systems because they resist fouling and offer excellent volumetric heat transfer. However, designing them involves a classic engineering trade-off between capital expenditure and operational efficiency.
A larger heat exchanger increases initial investment costs but reduces fluid velocities and pressure drops, thereby lowering the electricity required for pumps and fans. Conversely, a smaller unit is cheaper to manufacture but incurs high long-term operating costs due to increased pressure drops. This paper establishes a systematic economic optimization model to find the "sweet spot"—the physical dimensions and tube spacing that minimize the total present value of both capital and operational expenditures.
Methodology
The optimization framework integrates thermal-hydraulic modeling with lifecycle engineering economics.
Thermal-Hydraulic Model
The study utilizes established empirical correlations to model heat transfer and pressure drop. Specifically, Gnielinski's method is employed for calculating heat transfer coefficients, while Gaddis's equations determine pressure drops on both the tube side (water) and the shell side (air).
Cost Functions and Economic Parameters
The total cost is defined as the sum of the initial investment (heat exchanger materials, pump, and ventilator) and the present value of operating costs. Operating costs are calculated based on the electrical power required to overcome pressure drops, assuming 8,000 operating hours per year. Standard engineering economics formulas convert future electricity costs into present value, factoring in a nominal interest rate of 11% and an inflation rate of 7%. Equipment costs are derived by curve-fitting empirical price data from the 2016 unit price lists of the Turkish Ministry of Environment and Urbanization.
Economic Optimization and Key Insights
The core of the optimization is the visualization of how costs scale with the heat exchanger's physical dimensions, particularly its width ($B$).
As shown in the figure above, the initial investment costs ($FY$) scale linearly with width, while operating costs ($FI_{BD}$) decay exponentially as width increases and pressure drops fall. The summation of these curves ($FYI_{BD}$) reveals a distinct U-shaped profile with a calculable minimum total cost.
Another critical factor is the dimensionless transverse tube pitch ($s_T^*$). If tubes are packed too tightly to save space, the pressure drop (and thus fan/pump power) spikes dramatically, driving up lifecycle costs and negating any savings from a smaller physical footprint. The study found the optimal dimensionless transverse tube pitch to be approximately 1.55.
Results and Conclusion
The optimization model demonstrates that as the required heat transfer capacity ($\dot{Q}$) increases, both the optimal physical size ($B_{opt}$) and the minimum total cost ($F_{opt}$) scale linearly. For a baseline scenario of a 500 kW heat load using 3/4" steel pipes, the model successfully identifies the absolute minimum total present cost.
While the specific cost values are tied to 2016 economic data in Turkey, the underlying mathematical framework remains a robust tool for heat exchanger design. Future work could extend this model to finned-tube heat exchangers or incorporate variable-speed operational profiles to further refine the economic accuracy.
Summary updated