From: M—nica
To: Larry Bachus
Greetings from Colombia. I work in maintenance at a thermoelectric power plant. We are located about two hours from Bogotá. Earlier this year, our plant manager attended your pump seminar. He asked me to write to you regarding some pumps in my area.
We are developing a predictive-maintenance program on our centrifugal pumps. I hope to begin the program with a particular pump. See the attached factory performance curve (Figure 1). This pump was installed new eight years ago. We have changed 11 sets of bearings on this pump since the original installation. We install new mechanical seals with each bearing change. This pump rumbles and vibrates. We can’t locate the source of the vibrations.
We want to establish operating parameters on head, flow, bearing temperature, vibration, seal life, wear band gap, etc. to minimize maintenance. Your articles frequently mention the pump "sweet zone." How do you determine the pump’s sweet zone? Is discharge pressure and flow enough information to determine the pump’s actual performance?
Hello M—nica:
Many people in our industry use the term, BEP (Best Efficiency Point) to define the pump’s "sweet zone." Although the BEP exists, it is mostly impossible to operate a pump consistently at this point. System and operational variances drag the pump away from this point. A pump will oscillate through and around the BEP as the system changes, even in a correctly designed pumping system. It is more practical to consider a good-to-best efficiency zone. This is what I call the sweet zone.
At best, the engineer can design a pumping system so that the pump operator can manipulate a variety of devices including control valves, tank levels, pressure differentials, driver speed, and recirculation lines among others. These devices permit the operator to hold the pump within this sweet zone while the pump completes an operation.
The sweet zone is the area on the pump curve of best appropriation of energy in the form of head and flow. It is the zone of least wasted energy as vibration, heat, noise, and distortion.
Excessive vibrations and the accompanying noise and heat, will destroy the pump bearings, mechanical seals, and other strict tolerance parts on the pump. High heat will destroy the o-rings and other rubber seals and gaskets in the pump.
You may already know that most industrial pumps go into the shop with repetitive bearing and seal failure. Vibrations and heat lead to this unplanned, unexpected pump maintenance (the opposite of predictive maintenance). Pumps operated in the sweet zone are low-maintenance pumps, the result of good system design and operator training.
Next, you asked, "Is discharge pressure and flow enough information to determine the pump’s actual performance?"
You really need to know the differential pressure and flow (as reported on calibrated and functioning gauges and a flowmeter) to know the actual pump performance. The pump needs a suction pressure gauge and a discharge pressure gauge. In some cases a differential (PSID) gauge can work instead of two individual gauges. And you’ll want the pump curve handy too.
Reality suggests that most industrial pumps have no instrumentation and some pumps might have a discharge pressure gauge installed. The discharge gauge tells a production engineer that the pump is meeting its obligation to the operation, but says nothing about the pump’s health while it completes this duty. If the pump is unhealthy while performing, it becomes a high-maintenance pump.
It’s really simple — the pump takes suction head, adds energy, and produces discharge head. I’m saying the pump’s work starts at suction head. The suction and discharge heads can be converted to suction and discharge pressure gauge readings on your pump. As the engineer, you are really interested in the differential pressure across the pump.
The suction pressure gauge reading also indicates the energy available to the pump (the NPSHa). This must be greater than the energy required (NPSHr) by the pump (at its duty point on the curve) to prevent starving the pump.
Now let me offer some thoughts on the pump curve you attached in your message.
The curve you attached is the application specific curve for the 7.60-in. impeller at 3,500 RPM. The BEP is 80-81 percent at 560 GPM @ 220 ft. of head.
A set of coordinates are marked and labeled as the "design operation point." You (or someone) needed a pump to complete an operation. You gave a set of operating coordinates to a design engineer or the pump manufacturer. Based on your information, they reported that the pump should run at the design operation point on its curve (Figure 2).
I took your curve and marked coordinates that represent about 85 percent to 115 percent of the BEP. This pump is actually designed to run as much as possible between about 475 GPM @ 230-ft. of head, and about 640 GPM @ 210-ft. of head. This is the sweet zone of your pump. To be precise, the system should be designed so that the pump operator can hold the pump within these coordinates as much as possible.
You can see that the design operation point is outside the sweet zone. This is not necessarily bad if these coordinates represent the beginning, end, or any extreme (such as high resistance) of an operation. But you have the wrong pump if the design operation point represents the principle operating coordinates of this pump.
Most engineers would call the design operation point the principle duty point with the breadth of the sweet zone representing the extremes of the operation. The system should be designed so that the operator can hold the pump within the sweet zone as the levels rise and fall in the tanks, valves actuate, filters clog, and temperatures vary.
There are problems and implications if the real system design operation point is at 440 GPM at 235-ft. Any system alteration that increases the resistance (backpressure, pinched valve, new control device installed, or alteration to the piping arrangement) will move your pump further to the left on the curve, and the operator has no play in manipulating the pump toward the sweet zone again. I believe a variety of the previously mentioned operational and design alterations have conspired to drag your pump back to the actual duty point you indicate on your curve.
My advice, if you continue to use the pump that belongs to that performance curve, is that you adjust the resistance on the system and allow the pump to return to its sweet zone on the curve. Let the pump do what it was designed to do. The vibrations and rumbling should cease (unless they originate from another source).
Another option is to replace this pump with another pump that has best efficiency coordinates at your actual duty coordinates (180 GPM @ 235-ft of head), if indeed this is what your system requires.
I hope this helps,
Larry Bachus (a.k.a. The Pump Guy)
Larry Bachus, founder of pump services firm Bachus Company Inc., is a regular contributor to Flow Control magazine. He is a pump consultant, lecturer, and inventor based in Nashville, Tenn. Mr. Bachus is a member of ASME and lectures in both English and Spanish. He can be reached at [email protected] or 615 361-7295.
www.bachusinc.com
