Part 5: Factors affecting pumping plant efficiency and testing protocol
By Ed Butts, PE, CPI
Throughout this series we have concentrated on well water chemistry, potential impacts from corrosion or encrustation on water wells and pumps, and various methods available for well rehabilitation. In this installment, we’ll introduce the important concept of pump efficiency and many of the parallel methods used to test and determine the separate and combined efficiency of well pumps, their drivers, and means of power transmission.
Pumping Plant Efficiency
Efficiency has different connotations to different people, but never has the word had more validity and importance than in today’s rapid, energy-conscious world.
To the typical salesperson, efficiency means convincing you their product will deliver more of something for less of your money. To engineers, efficiency means producing some type of final result, whether in work or a product, with the least amount of effort and thus expense. Basically, efficiency is the difference between a theoretical or perfect result and the actual result from a task, effort, or process.
In our day-to-day world in the water well business, efficiency usually applies to some kind of electrical, mechanical, or chemical efficiency. Stated more simply: the net output of a system divided by the gross input into the same system.
Before we can address ways to improve the efficiency of a pumping plant, it is important we’re all on the same page as to what the term means. For the purposes of this column, we will limit our discussion to only potable groundwater pumping plants. The overall efficiency of a groundwater pumping plant is the product from these three distinct subclasses of components shown and outlined in Figure 1:
Overall pumping plant efficiency (%) =
Pump (%) × Driver (%) × Transmission (%)
Pump: This category includes the internal and external losses resulting in the overall efficiency within the pumping machine itself, including the hydraulic and mechanical losses through and surrounding the pump impeller, pump diffuser or case, shaft, bearings, packing or mechanical seals, and suction and discharge ports
Driver: Although the driver typically revolves around use of an electric motor, it can also consist of an internal combustion engine or turbine. This discussion will be limited to electric motors and engines.
Transmission: The transmission means include all operational and conveyance elements between the pump and driver or pump/motor and ground surface. This can include: the column assembly consisting of the column pipe, line shafting, and intermediate rubber/bronze bearings for all styles of vertical turbine pumps, along with the electrical switchgear and connecting wire for electrically driven VTP units or a right-angle gear or belt drive and engine, clutch, and driveline for engine-driven VTP units.
Submersible pumping installations include the electrical drop cable and the electrical switchgear and the column or drop pipe frictional head losses.
There has been some disagreement if the column/drop pipe frictional loss should actually be a component and element of the pump design. But since the friction loss is related strictly to the size and length of the riser pipe and not inherently with the pump’s internal design, I believe this type of loss is correctly classified as a portion of the overall transmission loss.
There will always be some measurable efficiency loss in each of the three categories when one of the components listed are used during system operation. The product of the individual efficiency of each of these three separate components multiplied together results in the overall efficiency of the pumping plant.
There are generally three types of efficiency losses associated with a typical pump bowl assembly (Figure 2):
(1) Loss from the internal hydraulic friction of the impeller (the impact from the fluid moving through the impeller, including velocity head).
(2) Loss from the external frictional losses (skin or disc friction) of the impeller (the impact from the impeller moving through the fluid).
(3) Combined hydraulic and mechanical losses of fluid flow through the suction and discharge ports, screens, pump casing or diffuser, any additional suction/discharge piping and valves and the power losses due to outboard and inboard bearings, shaft, and mechanical seal or packing friction.
The losses within a given pump include the hydraulic and mechanical friction and losses due to the movement of fluid through the pump and related components. This is not only dependent on the velocity of the fluid, but also highly dependent on the type of material used for the components and whether the surfaces are relatively smooth and free from resistance (machined or coated). For example, a material such as bronze or plastic will generally offer less resistance to fluid flow than a material made from cast or ductile iron due to the irregularities and exposed casting grains occurring on the interior and exterior surfaces of the iron parts.
For the driver, an electric motor loss includes efficiency losses due to the inherent internal electrical losses in the motor, beginning from the creation and maintenance of a rotating magnetic field, rotor inertia, hysteresis and power losses, and the thrust and radial bearing losses.
Losses in an internal combustion engine include those from various forces and transition from the conversion of fuel (i.e., chemical potential energy) to kinetic energy. The many losses include incomplete combustion, timing, parasitic losses (cooling pump and fan, alternator, fuel and oil pumps), air intake, and exhaust loss. Also included are radiation and frictional forces such as internal, inboard, and outboard bearing losses and internal losses across the various sleeves, pistons, rods, couplings, and other wear surfaces. All of this results in the primary loss in efficiency in an internal combustion power plant—heat.
Transmission losses include the loss due to the specific method of transmitting power from the upper driver down to the pump with VTPs or the power supply to the driver (motor) in the case of submersible units.
For a deep-well VTP, the transmission losses are comprised of the rotational HP loss in the lineshaft and bearings from the driver to the pump, the hydraulic friction loss in conveyance of the pumped fluid, plus the electrical cable voltage or HP drop from the power supply to the motor or a right-angle drive gearhead loss plus the loss in the driveline between the gearhead and engine or motor.
For a submersible pump, the transmission loss occurs from the electrical cable between the power supply to the submerged motor and the hydraulic friction loss occurring in the drop pipe from the pump to the surface. The cable loss is demonstrated in a total HP loss while the conveyance loss is in feet of friction. All of these are sources of energy losses.
Specific Issues for Well Pumps
The procedure used to determine the need for a pump rehabilitation is not much different than for a typical water well rehabilitation and can often be conducted at the same time. However, in the case of the well pump, the pump can be fully removed for a complete evaluation and repair where the well must be examined and rehabilitated in place.
In the case of the well pump, four factors create the most common causes of a decline in pump performance.
External physical causes
These are the causes resulting from the physical issues which are external to the pump itself, usually introduced from the well. This results in damage to the pump, generally between the pump’s running surfaces or within the pump impellers. This generally includes erosive wear from sand or silt in the production water, cascading water situations or insufficient submergence (causing vortexing), resulting in air entrainment and introduction into the pump suction or inadequate NPSHA. This can result in pump cavitation and severe erosion, especially in the pump’s first stage (lowest) impeller.
Internal physical causes
This specific issue results from internal wear factors within the pump itself such as metal fatigue or wear between running surfaces due to inadequate clearances, excessive wear between dissimilar materials, bent shaft or worn bearings, and normal/abnormal internal wear.
External and internal chemical causes
These closely mirror many of the chemical causes related to well issues, but in the case of a well pump the plugging that can affect performance can either be on the external surfaces, usually in the intake screen or riser pipe, or on the internal surfaces of the pump, including plugging due to iron, manganese, iron bacteria, calcium hardness, bacteria-forming slimes or a loss of metal integrity due to corrosion from low pH, electrolysis (use of dissimilar metals), or graphitization (cast iron corrosion). All of these can be within the pump itself or the riser pipe.
External system issues
In addition to the typical physical and chemical causes of pump performance decline, there can also be known or unknown system issues causing problems with pump performance. The most common of these include situations where the head is much lower or the required capacity is much higher than the intended design point for the pump, resulting in the pump running to the far right side of its curve. This can result in a horsepower increase or internal erosion due to cavitation from excessive NPSHR conditions or an excessive internal velocity.
Another less common situation is low voltage or a loss of power/frequency/speed to the motor/pump (more common with a variable frequency drive or engine). Severely low voltage, a loss of frequency as low as .25-1 hertz, or a fuel starvation issue with an engine can disrupt output speed and generated horsepower to the point where pump performance is also affected. This type of condition is much more prevalent where a VFD, generator, or engine is used to operate the well pump but can occur in ordinary electrical systems as well.
Finally, system issues can also involve changes in system design conditions from external modifications or additions within the system itself. Examples of this include inserting new obstructions (control valves, filters, and sand separators) into the system or discharge line, using smaller pipe than recommended for the flow rate (velocity > 5.0 FPS), modification of nozzle sizes or types of irrigation systems (changing a system from sprinkler irrigation to drip irrigation or vice versa), or erosion of nozzles from sand.
These can all result in a severe drop in pump and system efficiency, which is often not due to a specific loss in pump performance itself, but because of asking the pump to perform a duty it was not designed for. To either eliminate or factor this potential into the test results, all dynamics of the system should be evaluated for any relevant changes before undertaking repair or making conclusions from any pump system testing.
Pumping Plant Performance and Efficiency Testing
Although testing a pumping plant is often conducted in parallel with a well efficiency and performance test, it is not required and may present more obstacles to an accurate test result than desired. Remember proper testing of a well and the pump in the well require different observations and frequency. Where a complete testing procedure for a well may require many hours to perform, usually only a few minutes are needed once the well and pumping conditions have stabilized to obtain the needed data for a pumping plant test at each observation point.
Typically, a uniform set of testing and observational criteria should be adopted and used for all pump tests. Good examples of a proper test procedure and layout for a vertical turbine or submersible pump test are shown in Figures 3a and 3b.
In order to result in an accurate and meaningful test, all measurements must be taken only during and after stable well and pumping conditions and from representative locations. In other words, well static and pumping water levels should be obtained and referenced from an established elevation—such as the adjacent ground level and not from the top of the well casing or pump pedestal. Pressure readings should be obtained from accurate pressure gauges within the approximate 25%-75% span of the gauge, if possible. Pressure observation locations should be placed on the centerline of the discharge pipe and less than 1 to 2 feet from the discharge ell or head.
Obtaining accurate flow measurements are often more problematic and may consist of such wide-ranging elements as accurate as using a new inline flowmeter, orifice, venturi, or weir on the pump’s discharge, to as approximate and simplistic as an open pipe estimation of flow (Figure 4), to as elemental and easy as simply counting the number and verifying the diameter of sprinkler nozzles on irrigation systems.
Various types and formulas of flow estimation are available in most industry publications. Generally, the most accurate plant efficiency readings and results will be obtained from installations in which the pump and motor are designed for and operating at a fixed and full/single speed condition. This permits direct comparison to the original design conditions and operating curve and avoids the approximate corrections for multiple rotational speeds occurring with the affinity laws or changes from separate observations.
In addition, taking power or volt/amp readings from the incoming main power supply where multiple loads are present downstream will greatly distort the power readings and require correction for all loads unrelated to the pump and motor. When possible, for electrical installations all electrical observations should be taken as close to the motor as possible.
An example of this is at the motor controller load terminals or at the motor or wellhead (somewhere at or between the motor starter load terminals and motor). If the distance between the starter and motor is relatively short, around 20 feet or less, correction for the cable loss is not generally needed. But for long distance offsets or all submersible settings, the loss in the electrical cable becomes a significant factor and must be corrected to maintain accuracy.
The electrical loss associated with the drop cable in a submersible installation or the mechanical lineshaft friction and thrust bearing losses associated with a VTP cannot be overlooked or ignored. Each of these respective values are a component of the entire pumping plant’s efficiency and must be factored and included for accuracy.
The values shown in Table 1 and those in Table 2 can be used to approximate the carbon steel lineshaft or copper cable HP loss. However, I recommend always seeking out and using the actual data from the pump or motor manufacturer whenever available.
For a VTP: With 1.25-inch lineshaft at 1800 RPM:
From Table 1: .82 HP/C × 1.60 (for 160-foot lineshaft length including head and bowl shafts) = 1.312 HP loss.
For a submersible: 50 HP FLA = 68 A, cable selection #4 copper at 150-foot pump set plus 30-foot offset = 180 feet.
From Table 2: Cable HP loss for #4 copper conductor at 68 FLA = ~.54 HP/C × 1.8 (180′) = .972 HP loss.
Real World Example
In order to provide a more meaningful analysis of a pump’s wear, we have used the information observed and gained from a sample well test shown in Part 3 of this series.
1000 GPM at 32 psi discharge pressure at 80′ PWL = ~154′ plus ~5′ (riser hf) = 159′ TDH
600 GPM at 65 psi discharge pressure at 62′ PWL = ~212′ plus ~3′ (riser hf) = 215′ TDH
300 GPM at 89 psi discharge pressure at 54′ PWL = ~260′ plus ~1′ (riser hf) = 261′ TDH
The theoretical horsepower (water horsepower or WHP) required for each operating condition is:
1000 GPM: 1000 GPM × 159 feet TDH/3960 = 40.15 WHP
600 GPM: 600 GPM × 215 feet TDH/3960 = 32.575 WHP
300 GPM: 300 GPM × 261 feet TDH/3960 = 19.277 WHP
This well pump is equipped with a 60 HP, 1760 RPM, 3-phase electric motor. There is no model number or curve shown or available for the pump. In addition to the above performance data, the motor’s power input was also recorded at each capacity with the following results.
Method 1. Using Power Input from Watthour or Power Meter
1000 GPM at 159 feet TDH. Power input to motor = 53.19 kW (L1: 17.73 kW, L2: 17.73 kW, L3: 17.73 kW)
600 GPM at 215 feet TDH. Power input to motor = 40.37 kW (L1: 13.45 kW, L2: 13.46 kW, L3: 13.46 kW)
300 GPM at 261 feet TDH. Power input to motor = 35.00 kW (L1: 11.70 kW, L2: 11.85 kW, L3: 11.45 kW)
Method 2. Using Voltage and Amperage
1000 GPM: Average 3-leg motor amperage = 79 A. Average line to line voltage = 486 VAC
600 GPM: Average 3-leg motor amperage = 59.9 A. Average line to line voltage = 486 VAC
300 GPM: Average 3-leg motor amperage = 52 A. Average line to line voltage = 486 VAC
The power factor (PF) at each observation flow for both methods was 80% (.80)
Using Method 1: Use total power input in kW, convert to IHP (kW × 1.34), and divide into the theoretical water horsepower:
1000 GPM: IHP = 53.19 kW × 1.34 = 71.275 IHP. Plant efficiency (PE) = 40.15 WHP/71.275 IHP = 56.33%
600 GPM: IHP = 40.37 kW × 1.34 = 54.10 IHP. Plant efficiency (PE) = 32.575 WHP/54.10 IHP = 60.21%
300 GPM: IHP = 35.00 kW × 1.34 = 46.90 IHP. Plant efficiency (PE) = 19.277 WHP/46.90 IHP = 41.11%
Using Method 2: Use observed average amperage and voltage values from all collected readings into the following formula (for 3-phase) and plug the kW result into the Method 1 formula shown above:
= 53.20 kW × 1.34 =71.28 IHP. PE = 40.15/71.28 = 56.3%
= 40.34 kW × 1.34 = 54.05 IHP. PE= 32.57/54.05 = 60.26%
= 35.01 kW × 1.34 = 46.9 IHP. PE = 19.27/46.9 = 41.1%
In cases where a power meter is not available or individual electrical loads can be segregated on a single electrical service, a single motor’s input power consumption can usually be determined from the utility watthour meter by using the following formula:
Input HP (at meter) = 4.826 × K × M × Alt: kW (kilowatts) = IHP × .746 or 3.6 × K × M ×
K = Disc constant (The “K” value is typically shown on the meter face)
M = Meter multiplier (for installations with CTs). (Use 1 for no CTs)
R = Number of meter disc revolutions in T seconds
T = Time, in seconds for R number of disc revolutions
Example from above at 600 GPM “K” value = 13.45 R revolutions in T seconds = 20 revolutions in 24 seconds
(No current transformers (CTs) are used and meter is direct reporting, so M = 1)
Input (motor) HP = 4.826 × 13.45(“K”) × 1.0(M) × 20 revolutions (R) = 64.90 × .833 = 54.10 IHP
24 seconds (T)
Many current models and newer styles of watthour meters now use digital technology and report the load (usually in kW) directly in an LCD or LED format onto the meter front along with a rotation of other applicable data. When considering using this type of power monitoring method, always examine and record the watthour meter’s model, serial number, and site location and check with the serving utility to verify the capabilities or limitations of the specific meter and reporting units.
As you can see, the plant efficiency using both methods is roughly the same for each flow rate condition. This is based on the known relationship of an electric motor’s operating amperage and voltage (kVA) to the total power draw (kW).
In fact, the only difference between the kVA and kW is the power factor. Although the power factor is often considered a mysterious and difficult to determine electrical value, it is actually quite simple to find by using a clamp-on power meter. These meters are fairly inexpensive to rent or purchase and should be used whenever testing a pumping plant efficiency for installations using electrical motors.
The slight difference between the observed input HP using the input power readings in Method 1 and the volt/amp values in Method 2 could easily be caused from the low amount of power draw from the motor’s magnetic starter or other attached electrical devices such as relays, solenoids, or timers. Normally, the effect from these added electrical devices will be negligible or have little impact to the pumping plant’s overall efficiency.
However, in cases where other substantial or parasitic electrical loads are present in which the utility’s watthour or power meter will sense and record these loads, some correction will need to be made to prevent errant results. This is a primary reason I prefer when possible to use a separate power meter or the volt/amp measurements on the motor’s load conductors or terminals at the wellhead to determine the actual input horsepower draw from the motor itself.
From my observations, the two most common errors in determining the pumping plant efficiency of a particular installation are:
- Not using or maintaining the same units throughout a measurement or calculation
- Not using or estimating an accurate power factor.
The use of uniform values throughout calculations is critical to the final determination of a plant efficiency.
Where a conversion from kW to HP or change from psi to feet of head is required, it is vital the proper correction multipliers are used throughout.
Regarding the power factor: If it is not possible to determine the precise power factor, an estimation of this value can be used for most electrical installations. Depending on the distance from generating and distribution and transmission facilities, I often use an estimated value of .80 (80%) to .90 (90%) for a system’s power factor. Once again, the key to using this procedure is uniformity. It is critical the same estimate is used for all calculations since any variations will obviously impact the input load estimates.
The disparity of plant efficiency at the three separate flow rates is common for many pumps since the
pump efficiency typically varies as it moves away in either direction from the best efficiency point (BEP).
The plant efficiency of 60.2% demonstrated at 600 GPM and 56.3% at 1000 GPM along with the severe drop-off at 300 GPM indicates this bowl probably had a BEP lying somewhere between 600-800 GPM and efficiency beginning to decline as the flow moves in either direction away from 600 GPM. It can also mean the impellers’ wear rings or eye surfaces are severely worn and considerable eye or stage recirculation is occurring at the higher head values.
This type of plant efficiency test provides an overview of the efficiency for the entire plant and allows us to make a few educated guesses that the bowl assembly is at least 10 inches in diameter, given the HP, capacity, and pump speed, but does not necessarily indicate which component is high or low in efficiency. That kind of analyses requires a more detailed examination, including confirmation of the bowl diameter and number of stages, the size and length of lineshaft, and depth of pump setting.
But even without this information, we can surmise the plant efficiency, at 60%, is somewhat lower than desired for a typical 60 HP VTP electric driven unit. Generally, even with an older pump and motor combination we would hope to see a plant efficiency between 64%-66% for a 10-inch-diameter bowl at the pump’s BEP and up to 75%-80% for a new plant.
Engine-Drive or VFD Installations
Where a variable frequency drive (VFD) or engine-driven pump is used, the most critical observation is often determining the pump and motor (driver) rotational speed (RPM). Obtaining accurate speed data of the pump permits plotting the current pump performance against the original pump curve to find out any drop of performance. If the original pump curve or data is unknown or not available, this initial test will permit plotting of a new performance curve for future reference and comparison.
With a VFD the motor and thus the pump speed can be determined from the direct relationship between motor frequency and speed. For example, on a 60-hertz power supply, a readout that indicates an output frequency of 45 hertz represents a corresponding motor speed of 75% (45 hz/60 hz) of the motor’s nameplate speed. This means a 3450 RPM submersible motor will now be turning at ~2,588 RPM (3450 RPM × .75) and a 1760 RPM vertical lineshaft pump will be rotating at ~1320 RPM (1760 RPM × .75).
For a vertical turbine pump, whether electric or engine driven, a hand-held tachometer placed on the top of the head shaft in the driver will provide a true vertical pump speed regardless of any impact caused from a gearhead speed reduction/increase or a VFD. Use of the affinity laws can provide an approximation of performance at various speeds.
An accurate determination of the energy usage for an engine-driven installation can only be provided from the fuel BTU and consumption value. This can be conducted in various ways, but I have found the easiest method is to simply gauge and time the fuel consumption from a measured container or fuel tank or through the use of an inline fuel meter or gauge. Once again, since the shaft power and therefore fuel consumption will vary depending on the engine speed and power output, accurate monitoring of the engine’s operating performance, including the engine’s output speed and fuel consumption, must be tracked and compared against the actual pumping conditions that exist at the time of measurement.
Where a right-angle drive gear head (gear drive) is used to transfer the horizontal power from the engine to the vertical pump shaft, a knowledge of the gear drive’s speed ratio, up or down, and efficiency must also be factored into any meaningful plant testing.
I have included Figure 5, which can be used as a test-ready sheet and guideline to use in creating your own format for pumping plant efficiency testing procedures and calculations for most applications
This concludes this discussion on pumping plant efficiency and testing procedures. Next month, in our final installment, Part 6, we will wrap up this topic with an overview of the various current and new methods used to raise pumping plant efficiency for deep well pumps and drivers.
Until then, work safe and smart.
Ed Butts, PE, CPI, is the chief engineer at 4B Engineering & Consulting, Salem, Oregon. He has more than 40 years of experience in the water well business, specializing in engineering and business management. He can be reached at email@example.com.