What is pump system optimization? Is this another catchphrase similar to the word “quality,” or does it have a specific meaning? Pump system optimization should be known as a service that is both a science and an art form focused on maximizing pump system efficiency. While maximizing mechanical efficiency may not always be achievable or advisable, it should be the operative word.
Maximizing efficiency solves a multitude of issues in terms of minimizing electrical energy consumption in both fixed and variable speed systems while maximizing reliability. By the way, variable speed drives, with few exceptions, are much better process controllers than control valves in low static head systems. Furthermore, fixed speed systems that produce smooth flow near the best efficiency point (BEP) are easier to control with a modulating valve.
So how has it come of age? Understanding a pump in the context of a system has been a staple of pump training, selection and sizing since the beginning—or has it? We have had to know the process parameters that surround a pump, such as net positive suction head available (NPSHa), specific gravity, static head and design point, among others, in order to pick the right pump for application to the system. Also, pump specifications typically require that the selected pump has the highest mechanical efficiency when compared to alternate models.
But the key difference is when the specification does not require the pump to operate near the BEP once it is installed in the process. We select a pump with high mechanical efficiency, but then oversize the pump to ensure a safety margin before installing in the process. That’s why, according to statistically valid studies, the average pump system efficiency across all industrial market segments is 40 percent. Furthermore, these same studies show that 10 percent of centrifugal pumps operate around 10 percent efficiency—a massive efficiency loss.
Additional evidence of this condition is that process control audits focused on valve tuning and process variability find that 25 to 60 percent of control loops increase process variability and, in order to settle the process, are subsequently switched into manual. According to Process-Industry-Practice (PIP), control valves should be sized to be more open than closed at normal flow—for example, operate between 50 and 80 percent open—while actual industry practice often targets the valve to be 50 percent open at normal flow. Oversizing then results in control valve settings in the range of 10 to 50 percent open. As a result, our knowledge and practice of pump systems often leads to excessive throttling and inordinately low efficiency.
Improving System Efficiency
Whether intended or not, pump system optimization has come of age as it is no longer practical to leave industrial pumps in their current state. While environmental concerns have mounted since the 1970s, relatively little action has been taken in the interim to mechanical efficiency to reduce electrical demand and, thereby, reduce greenhouse emissions. That has been primarily due to the physical limitation of mechanical design.
However, a pump is a component of a system and, as such, its efficiency is largely dictated by process design, sizing and control scheme. Although there is a federal regulatory movement under way to improve overall mechanical designs of pumps through new component and material science, the best way to make quantum leaps in pump system efficiency is to make mechanical and control modifications to the system that ensure the pump operates closer to BEP under normal operating conditions.
Removing the specialized equipment that dominates the terrain of most industrial facilities—such as the catalytic crackers and distillation columns in a refinery or the paper machine in a mill—leaves the vessels, pipes, pumps, valves, instruments and controllers. And, almost unbelievably, there is little real-time information about the pumps and other rotating assets in the plant information and control systems.
Typically, at the distributed control center (DCS) or programmable logic controller (PLC), we know that the pump is on or off—that’s it, except for the occasional amperage reading from the motor or a bearing vibration signal that’s been hardwired into the system. In some cases, there are tens, even hundreds, of millions of dollars invested in rotating assets and infrastructure with the only continuous process readings being pipe flow, pressure, temperature, and tank level. Interestingly, a high percentage of pumps, estimated to be around 60 percent, have no flow meter installed either in the discharge line or near the end-user system.
The computerized maintenance management systems also have little to no real-time information, with repair records and cost-related data entered manually—often in an inconsistent matter. It is no wonder that maintenance costs typically exceed energy costs in manufacturing and often by an order of magnitude.
Times Have Changed
Certainly, if you can’t measure key performance indicators in real time, you can’t control in real time. That means you either run to failure or utilize preventive component replacement schedules as the primary options. Predicting when it will fail is the best option when the necessary measurement tools are available.
As described above, this is the state of our underlying mechanical systems in the process industries—and there is a good, rational reason that it has evolved this way. But times have changed and the knowledge, technology and market drivers are available to change trajectories.
First, we have to use the static and dynamic data that exist either in the field or at the process control and plant information systems in order to start determining pump system efficiency. Next, we have to develop and implement continuous-improvement plans for existing and newly installed pump systems. This involves programs to ensure these systems are optimized over future years.
If we really want sustainable processes, not just rhetoric, this has to be done starting now and extending into the foreseeable future. The time has come to accelerate pump system optimization, and the journey can no longer be delayed.