Figure 1. Pump curve for 5,000 gpm at 75 ft of head, resulting in a 125-hp motor.

While opportunities to improve motor and drive efficiency continue, pumps and fans remain overlooked candidates for significant savings through smart design. From the elevation of pump discharge to duct configuration on the fan inlet, the author offers multiple targets for smarter design that you may not have considered.

Reduced energy usage is the goal of most owners for their buildings and other facilities that are designed today. Toward this goal, various manufacturers have made design modifications of the motors and their driven loads over the last thirty years to reduce their energy consumption.

However, when a system has been engineered with basic components that have high losses, it is irrelevant whether the motor or the driven device is efficient, since the total energy usage of the system will be significantly higher than required.

One of these energy-saving efforts has been to reduce the energy usage of the electric motors that drive mechanical loads. Historically, electrical motors were in the range of 80% to 85% efficient in converting electrical energy to mechanical motion. For typical, small, integral-hp motors on the market today the general purpose motor is around 86%, while the high-efficiency motors approach 92%. Larger motors show a slight improvement in efficiency and exceed 95% for motors over 100 hp.

To put these numbers in a better perspective, the cost of operating a 7.5 hp, standard efficiency motor at full load, twenty-four hours a day, would cost about $8,000 a year (with an electrical cost of $0.10/kWh). Under the same operating conditions, a high-efficiency motor would cost about $7,400, providing an ROI of between one and two years for buying the more efficient motor.

Figure 2. Pump curve for 5,000 gpm at 75 ft of head, resulting in a 125-hp motor. Note that only the impeller was changed slightly using the same pump operating at the same RPM under the same conditions, to achieve a 25 hp reduction in motor size.

The selection of the motor model can have a dramatic impact on the operating efficiency. In a DOE study from 1993, the efficiencies of a 20-hp motor available at that time ranged from 86% to 93.3% for the various manufacturers studied, and these values are not significantly different today. While there have been some incremental improvements, every additional increase in motor efficiency will now come at a higher premium, so how do we reduce facility energy usage?


With a goal of improving the energy efficiency of a facility, instead of specifying motors with higher and higher efficiencies, perhaps a decrease in the losses in the piping systems could have a significantly greater impact? For example, a brief survey of water pumps showed a range of efficiencies from 60% to 90%, according to the specific pump, volume, and head. It should be obvious that selecting a pump with the same lift, volume, and hp characteristics, but with a higher efficiency, can only reduce the energy consumption. Even a 3% or 4% improvement can result in significant energy savings for the owner.

However, with further research, it was clear that much greater gains may be achieved by reducing system pump head! (As a part of this research, the math behind these dramatic changes in pump head became evident due to the relationship between the friction head vs. the diameter is the inverse fifth power.)

Taking a sample pump curve for a 5,000 gpm system with 75 ft of pump head as shown in Figure 1, (Bell & Gossett (B&G)10x12x13 1/2A) the efficiencies range from 60% to 85% and the hp from 100 to 300. For the assumed conditions, a 150-hp pump would operate at 77% efficiency. When the desired pump delivery volume is fixed, as in most applications, there are only two variables left: change to another pump with a different performance curve, or reduce the total pumping head. Changing to another pump curve may have some benefit, since going to a B&G 10x12x11 1/2 B (Figure 2) requires a 125 hp pump operating at 79% efficiency. But reducing the total pump head has the potential of much greater energy savings than any alternative pump or energy conservative motor.

Pump head is a function of water velocity, piping resistance, and vertical lift. Vertical lift, in many applications, is fixed, but where there are options in modifying the elevation of the pump discharge or mounting location, there is a resultant reduction in the pump’s hp requirements and, thus, its energy usage. If the pump operating curve is just above a hp break, changing the lift by five or ten feet can reduce the pump head enough to reduce the pump’s hp to the next smaller size.

Table 1. Frictional losses in pipe fittings.

While the efficiency changes very little (whether up or down), using a 100-hp motor operating at 60% efficiency utilizes much less energy than operating a 150-hp motor at a higher efficiency. For example, reducing the pump head from 80 to 55 ft changes the operating point for the 10x12x11 1/2 B to a 100-hp motor.

Noting the various relative values of pipe-fitting resistances shown in Table 1, there are a number of opportunities of reducing the frictional resistance of the pipe run. Utilizing long sweep elbows in place of standard elbows will result in a reduction of 30% to 35% of the losses for each elbow. Also, requiring the contractor to keep the number of fittings to an absolute minimum will also keep the piping losses to a minimum. However, noting the losses in the three types of valves, a reduction in the valve losses, or the selection of valves with lower losses can dramatically reduce the pressure drop. For some installations, there may be no choice on valve selection, so other means should be considered.

Table 2. Frictional losses in straight pipe.

In the example in Table 2, the same number of fittings were used, and only the pipe diameter was changed. For a flow rate of 1,000 gpm, pipe sizes of 5, 6, 8, and 10 in. were utilized in calculating total pumping head. While most engineers would not consider 5 or 6 in. pipe for this application, they might consider using an 8-in. pipe. The reduction in head losses in increasing the pipe size from 8 to 10 in., resulted in being able to reduce the pump motor by 2.5 hp. The reduction of the motor from 10 to 7.5 hp resulted in an annual savings of $830 or a savings of over $20,000 for the 25-year life of the pump.

As previously noted, the reduction in vertical head had the most impact on the pump operating hp. In many cases, the vertical lift of a pump cannot easily be changed, but for the cases where it is possible, the hp of our example pump frequently dropped by more than 50%.


There are similar opportunities for energy savings when selecting supply/exhaust fans. The four fans, which all deliver 30,000 cfm at 5 in. of static pressure, are manufactured by the same company and are of the same class, just different blade configurations, widths, and RPMs (Figure 3). Note that the four fans produce the volume of air at the same conditions, but require horsepowers of 10, 27, 33, and 48. Selecting the fan with the lowest hp requirement will, obviously, result in significant operational savings due to less electrical energy usage. Other styles of fans by this same company or other companies with different blade designs and dimensions could still supply the same 30,000 cfm at 5 in. of static pressure and may require a motor of only 25 hp, or they may need as much as 600 hp. Selecting the fan that can supply the required volume and most efficienctly pressure can result in significant operational savings for the facility owner.

Air supply fan operation is affected by the design of the duct work, so lowering the system static pressure will allow the use of a smaller the fan motor. Pressure drops in straight runs may be reduced by increasing the cross-sectional area of the duct. For large, open ceiling spaces, that would work, but having lots of space for duct work in the ceiling is quite rare, so this method of reducing static pressure is not as useful as some others.

As in piping, the elbows or turns are another opportunity to reduce static pressure. A square turn in the duct work, with no inside or outside radius, has a static pressure loss that may be as much as 100 times the loss of the same turn with radiused interior and exterior and turning vanes. Likewise, square transitions from one duct size to another will have significantly greater static pressure losses than ducts that utilize very gradual transitions.

Where multiple diffusers are fed from a single main duct run (the most common supply duct configuration), there are many opportunities for reducing pressure drop. The highest pressure drop for a diffuser take off is the design in which the take-off is perpendicular to the main duct run. Designing these take-offs as diverging wyes with interior extractor vanes, instead of a square duct turn, will result in dramatic reductions in the pressure loss for each diffuser branch. By combining the wye take-offs with static regain sizing of the duct runs and carefully calculating the appropriate losses in each branch, balancing dampers (one of the greatest sources of static pressure gains in duct runs) can be virtually eliminated. At worst, the dampers that are integral with the diffusers might be all that are required for balancing a system.

One of the most overlooked design issues that can cause a major increase in fan static pressure is the inlet/outlet conditions of the fan. When fittings are located near the fan, the effects on static pressure are much greater than when they are further away. The distance that a straight duct should run - prior to making any turns, take-offs, or size changes - is 2.5 times the effective duct diameter for velocities of 2,500 fpm, increasing to six times the diameter for velocities over 6,000 fpm. The fitting location on the fan discharge is critical so the air can achieve laminar flow once it enters the duct run.

Oddly, the duct configuration on the fan inlet is just as critical or even more so than the discharge. For squirrel cage-type fans with the air inlet at the fan’s axis of rotation, utilizing a rectangular box duct with the entrance perpendicular to the fan axis, has been observed to reduce the fan capacity by as much as 45%. This inlet condition is so detrimental to the fan operation that no methodology has been developed to calculate the exact losses, so only observation data are available.

By implementing every option to decrease the static pressure in duct runs, fan hp can be reduced considerably. Expanding on the prior example of the 30,000 cfm fan, the same fans were re-investigated, using the same air volume and reducing the static pressure operating conditions, there were dramatic reductions in fan hp. A decrease of the static pressure from five to four inches can reduce the fan size from 50 to 40 or even 30 hp, according to the specific fan selection. With additional reductions in the static pressure and maintaining the air volume at 30,000 cfm, a very low static pressure of 0.5 in. of static requires a motor of just over 10 hp and would warrant selection of a smaller fan.

Obviously, many installations will not lend themselves to a ten-fold reduction in static pressure, but every incremental pressure reduction will result in a corresponding reduction in fan hp. Even when the reduction is so small that one cannot select a lower hp motor, the installed motor will operate at a lower input current, thus saving energy for every hour of operation.


The energy conservation laws and codes that have been enacted in the last 35 years have made a significant impact in the total energy utilization of a facility. From these brief examples, one can see that the design engineer can influence the energy requirements of a building as much or more than any of the conservation measures contained in these laws and codes. By engineering a design that has the lowest pump head or static pressure possible for that particular system, the operating cost may be reduced dramatically.

With the inclusion of very simple changes in component design and placement, the return on the first-cost investment during construction, could have a pay back of one to two years. With all of the research and debate on achieving a carbon neutral society, reducing the losses in these systems will go a long way toward reaching this goal.

Conversely, using an energy-efficient motor with a well-designed pump or fan is completely wasted (e.g., equaling or increasing the carbon footprint for this system) when combined with an inefficient system design.ES

Sidebar: Pump Or Fan Substitutions

Many times during the construction of a facility, an engineer receives a request from the contractor to substitute an alternative manufacturer of a pump or fan. These substitutions should be carefully reviewed by the engineer, and the impacts of the long-term energy usage should be determined prior to acceptance.

For example, a 40-hp fan might have a delivered air volume of 30,000 cfm at a 5 in. static pressure. The contractor proposes that he will furnish a 50-hp fan that has exactly the same air volume, static pressure, and efficiency characteristics, but it requires a larger motor to achieve this performance. If the fan operates at 80% full load, sixteen hours a day, for an average work year of 260 days, the difference in operating cost is $3,600 per year. For a fan with a 25-year expected life, this amounts to $90,000 with no factor included for inflation. This is significantly greater than any small savings which the contractor may offer in a contract amount reduction!