Pumps & Systems, September 2007
Back in the old days, level control had little or nothing to do with saving energy. In fact, it was often a necessary evil. Today, that is no longer true - the VFD offers the potential for power savings in lift station applications that range from a few hundred gallons per minute to those that have to move thousands of gallons each minute.
A Little History
In an ideal world, the flow of wastewater into a wet well would be constant. We could then size a pump (or pumps) that could run continuously and remove that entire inflow at the same rate as its entry. The result: smaller pumps that run at BEP 24/7. We learned early on that this approach would be impractical unless we could gain total control over the habits of the populace. Although this has been attempted in other parts of the world, history has taught us that it seldom works. These revelations led to a less desirable but more practical alternative known as "pump down."
Pump down is extremely simple - when the water in the wet well rises to some maximum level, a pump starts and pumps the well down to some predetermined lower level. The pump then shuts down and waits for the water to rise again. Usually a wet well is sized for some minimum pump run time in order to keep the number of pump starts within the guidelines of the manufacturer. Some may be oversized and employ multiple smaller pumps in an attempt to emulate that Utopian system we envisioned originally. Overall, pump down can be a very effective and relatively efficient process.
Historically, one of the challenges of lift station design, especially high flow ones, has been keeping pump starts to an acceptable level. In some cases, this can be attained by installing multiple pumps and alternating them with each successive pump down cycle. Another method is to stage multiple smaller pumps and attempt balance outflow with inflow (level control). Although both of these methods work well in many installations, there are times when the necessary wet well volume becomes unrealistic or the number of staged pumps required cannot be accommodated.
An alternative approach would be to vary the pumping rate by changing pump speed. This would allow outflow to be closely matched to inflow, and thus significantly reduce the number of pump starts.
In the 1960s, well before the arrival of the VFD, a variable speed technique became available that employed a motor originally developed for constant torque applications.
Rather than using the imbedded bars found in a standard rotor, this rotor is wound with coils of insulated copper wire that terminate at a set of slip rings. Brushes, similar to those found in DC motors, allow a resistance to be connected to the coils during motor operation. By varying the resistance of the rotor coils, the slip speed of the rotor can be altered and that range varies from rated speed when the coils are shunted to approximately half speed at some maximum resistance.
The device that was used to vary the resistance seen by the rotor was known as a "liquid rheostat." Rube Goldberg died in 1970, but I think he must have spent his later years helping develop this machine. It consisted of a large tank that contained an electrolyte (salt) solution and movable metal rods that were wired to the rotor brushes.
When the rods were immersed deeper into the solution, resistance was lowered due to increased surface area contact, and motor speed increased. When they were raised, resistance increased and speed decreased. A bubbler system determined the wet well level and controlled a linear motor that raised and lowered the rods in an attempt to keep the level constant. The heat that was generated by the increased resistance was removed from the electrolyte solution by a circulating pump and heat exchanger.
Rube would have been proud because it was a complex contraption and required considerable maintenance, but it met its goal: dynamic level control and decreased pump starts. The liquid rheostat was popular in the wastewater industry throughout the 1970s and 1980s, and was still in use in the early 1990s.
VFD Level Control
The liquid rheostat was all about controlling level for the purpose of reducing motor starts. Even though the required HP was reduced at lower speeds, the heat generated by the higher resistance tended to cancel any potential energy savings. Also, in order to keep from stopping the pump, it would often be forced to run at lower than optimal speeds, which resulted in low hydraulic efficiency.
Today, we can achieve level control without these drawbacks and save energy to boot. Let's compare a pump down and level control application that uses the same pump.
Figure 2 shows the performance of a 12-in non clog with a manufacturer approved flow range of 800-gpm to 5500-gpm in a pump down application.
The red horizontal line at 22.5-ft is the static head seen by the pump when the wet well is full, and the brown line at 32-ft is the maximum head at the end of the pump down cycle. The cycle begins at 4500-gpm and ends at 3000-gpm, and hydraulic efficiency is maintained at a healthy 77 percent to 78 percent across the entire range of the pump down cycle.
At first glance this appears to be a pretty efficient operation. All pumping occurs at or very near BEP, and HP drops as we approach the end of the cycle. But if we divide the HP required at each major flow point by the flow in GPM at that point, we will gain an entirely different perspective.