Users of heat processing equipment are acutely aware that the energy consumption of the equipment is a significant cost. Nationally, the price of electricity for industrial users increased 15.7 percent from 2004 to 2006. This trend is pushing equipment users and manufacturers to look for new and innovative ways to reduce energy consumption. One method where remarkable results can be achieved without significant increases in capital cost is to use computational fluid dynamic (CFD) analysis to optimize the flow of air through equipment.

The electricity required to drive a fan depends directly on the volume moved and the pressure drop through the system. Often, the volume of air is critical to the process and cannot be reduced. The pressure drop, by contrast, generally has no direct benefit to the process.

Pressure drop depends on the square of velocity; therefore, the easiest way to reduce pressure drop is to reduce velocity. If the volume is dictated by the process, this means making the air passages larger. This adds cost to the equipment and often runs up against practical limitations on the size of the machine or available space.

In many cases, the system geometry causes uneven flow distribution. An increase in pressure drop and turbulence will occur and increase in magnitude as the flow-field becomes more non-uniform. Simple modifications to geometry and the informed use of turning vanes can improve flow uniformity, and thus lower pressure drop and operating cost.

The following studies show how CFD analysis can be used to reduce pressure drop and do so while maintaining or improving process integrity. Remarkably, the capital costs of the resulting designs are the same or less than the original design.

Fan Discharge Transition

Enlarge this picture Figure 1. The duct connects the fan discharge to the dryer plenum. It was redesigned based on the results CFD analysis.

Figure 1a shows a section of duct that connects the discharge of a fan to a plenum in a dryer. Air enters the top of the elbow on the left (shown in brown) and exits the right-hand side of the duct transition (shown in green). The exit of the duct section has a larger height than the entrance. Traditional airflow duct design would suggest that increasing the height of the transition will lower the air velocity and decrease the pressure drop. However, CFD analysis shows that because of the non-uniform velocity discharge profile coming from the fan, the vertical flare in the duct actually causes the flow to be less uniform, which increases the pressure drop (figure 1b).

Based on the results of the CFD analysis, the transition duct was redesigned to extend the turning vanes in the elbow and keep a constant vertical height in the duct. The new geometry (figure 1c) and corresponding CFD analysis (figure 1d) show that by making this change, the resulting pressure drop changed from 0.95" water column (w.c.) to 0.82" w.c. The revised duct will be simpler to fabricate, use less material, and take up less space.

Supply Plenum

Enlarge this picture Figure 2. In the supply plenum, air enters the back left sidewall of the box below the green plane, a perforated plate through which the air passes in a vertically upward direction. CFD analysis suggested a design that achieved a pressure drop of 1.30" w.c.

In the second example, a supply plenum, air enters the back left sidewall of the box below the green plane (figure 2a). The green plane is a perforated plate through which the air passes in a vertically upward direction. The process occurs on top of this plane, so air distribution exiting the perforated plate is critical.

The momentum of the air entering the plenum creates a natural tendency for greater discharge on the opposite side of the plenum. Figure 2b shows the distribution of airflow exiting the top of a plenum, which is uniformly 23 percent open area. The pressure drop is 1.49" w.c.

Reducing the open area of the top of the plenum increases the pressure drop, which makes the air distribution more uniform. A traditional approach would be to use a very low percentage of open area (and a very high pressure drop) to overcome the problematic geometry and make the airflow more uniform. Figure 2c shows the air distribution out the top of the plenum with a uniform 5 percent open area. While the distribution is now very uniform, it comes at a steep price. The pressure drop increased from 1.49" w.c. to 8.10" w.c., a dramatic increase in pressure drop and energy consumption.

An alternate approach is to use CFD to determine variable open areas that will overcome the inherent geometric problems without leading to greatly increased pressure requirements. Figure 2d shows variable open area layout for the top of the plenum, and Figure 2e show the resulting flow distribution. The distribution is satisfactory for this process and is achieved at a pressure drop of 1.30" w.c. Note that the average open area is about 20 percent; yet, the pressure drop is less than with the uniform 23 percent open area because the flow is much more uniform. This again illustrates the point that non-uniformity causes higher pressure losses.

Editor's Note: A third exampled of how CFD can help optimize oven and dryer design is available by using the link to a web-exclusive sidebar at the bottom of the page.

The Payoff

In conclusion, the examples are all part of a dryer that operates in a process that calls for 78,000 acfm airflow. The original system design has a total pressure drop of 6.58" w.c. and very poor air uniformity exiting the 23 percent open perforated plate. A typical fan with backward-inclined blades would require 149 hp in this application and would have a 200 hp motor. Once the modifications (including the return duct change described at www.process-heating.com) were made, the system design had a pressure drop of 3.37" w.c. and much improved air uniformity and process performance. The same fan now requires only 112 hp and would have a 125 hp motor.

The change results in significant savings in operating costs associated with electricity. The national average electric price for industrial users was $0.0610 per kilowatt hour in 2006; based on electric price trends, a reasonable assumption for 2007 is $0.0630 per kilowatt hour. A 37-hp reduction will save approximately 29 kW. For a 2,000-hour-per-year operation, this saves $3,650 per year per fan. The bigger the scale, the bigger the savings will be. For a facility that operates 24 hours a day, 7 days a week, and 50 weeks per year, the savings is $15,350 per year per fan. Now consider that a piece of processing equipment may have several fans, and the savings can grow to more than $100,000 annually.

Furthermore, the cost savings for the smaller motor, motor starter, wiring and smaller supply duct more than offset the added fabrication costs associated with the turning vanes, duct transition and variable-perforated plate pattern. The new design lowers capital cost, lowers operating cost, and improves the process.

The software required to perform this type of analysis is available from several vendors. The accuracy of the outputs is limited only by the accuracy of the inputs. The success of the CFD analysis is greatest when the software is benchmarked and an experienced user is doing the study. Otherwise, the potential exists for the analysis to be a case of garbage in, garbage out. When applied correctly, however, the result is data in, money out.