In 1997, a leading American carpet and textile manufacturer was building a plant in Shanghai. The original design, done by a leading European design firm, required 14 pumps for a runaround heat transfer loop and required 70.8 kW_e of total power. But Jan Schilham of Interface/Holland took a fresh look at the design and reduced the total pumping power of the heat transfer loop to only 9.7 kW_e, 86 percent less, with lower capital cost and better performance—thanks to two changes in design mentality.

First, Schilham chose to use fatter pipes and smaller pumps rather than the specified small pipes and big pumps. Since friction is inversely proportional to (approximately) the fifth power of pipe diameter, making pipes 50 percent fatter will reduce friction by nearly 86 percent. Pump size (and roughly cost) will fall proportionally with the reduction in friction. The capital cost of the pipe is roughly proportional to the second power of pipe diameter. So clearly it is better to use fat pipes and small pumps. But why weren’t the bigger pipes selected the first time? Traditionally pipe size is optimized against only the pumping energy cost, and pipefitters ignore the size—and capital cost—of the pumping equipment. Optimizing the whole system—pumping energy plus capital cost savings—yielded fat pipes, tiny pumps, and lower capital and operating costs.

Second, Schilham laid out the pipes first, and then located the equipment they connect—the opposite of how systems are typically installed. Typical pipe runs twist and turn to hook up equipment that’s far apart, separated by extraneous stuff, facing the wrong way, and mounted at the wrong height. This raises friction by about three- to sixfold—delighting pipefitters, who are paid by the hour, mark up the extra pipes and fittings, and don’t pay for the bigger pumping equipment or electric bills. By making the pipes short and straight, the pumps, motors, and electrical components could be made even smaller and cheaper.

In addition to lower capital cost and a sevenfold lower pumping power requirement, the redesign also yielded additional free benefits, including 70 kilowatts less heat loss via easier insulation of short, straight pipes. Other bonuses included simpler and faster construction, smaller floorspace and weight, easier maintenance access but less need for it, higher uptime, and longer life as a result of fewer erodable elbows.

Another important general lesson to learn from this case is that the right steps need to be done in the right order. If larger pumps were selected first, and then the pipes were optimally selected and arranged, the pumps would be oversized, and the system would be inefficient. Doing things in the right order can maximize the favorable interactions between components.