In 2014, this column focused on improving the operation of piping systems by performing pumped system assessments. The columns have covered energy cost balance sheets, the assessment process and the performance of sample assessments. By now, readers are aware that all the equipment found in a piping system falls into one of three categories:

  • Pump elements add all the fluid energy to the piping systems.
  • Process elements are used to make the product or provide the service.
  • Control elements are used to improve the quality of the product or service.

The next series of columns will focus on how these basic elements are used within the piping system and how they work together to achieve system operational objectives. With an established understanding of the equipment used, operators can evaluate the total system, which leads to a better understanding of what can be done to reduce operating, maintenance and capital costs while improving system uptime.

Pump elements include the pump, the drive and the ancillary equipment needed for pump operation. Many types of pumps and drives are available, but for the purpose of this column, the discussion will be limited to centrifugal (or rotodynamic) pumps driven by alternating-current induction motors. These are the most common pump and drive combinations used in industrial piping systems.

Why Pumps Are Typically Oversized

A common scenario often leads to incorrectly sizing pumps. In this scenario, a company decides to release a new product. The results of its marketing study project sales of 50,000 tons of product per year for the first two years. After this, sales would increase to 100,000 tons per year. An annual sale of 200,000 tons per year is projected after five years.

The engineering, procurement and construction firm (EPC) is contracted to design and build a new plant and ultimately allow for the production of the product. During the discussions between the client and the EPC firm, a mood of optimism prevails. The plant is designed to meet the future estimated capacity of 200,000 tons per year. The engineering group determines that the processes used to make the product needs a series of simulation studies to optimize the design. This includes the required mass flow rates, fluid physical properties, and the pressure and temperature requirements for each of the plant’s systems.

The mechanical engineering team uses the process flow diagrams provided by the process engineering group to size the equipment needed. The team determined the following:

The example system requires a pump with a capacity of 800 gallons per minute (gpm) to achieve the plant’s design objectives (see Figure 1).

Figure 1. This single system is one of the many interrelated systems designed for the total process. (Graphics courtesy of the author)

During the preliminary design process, the long lead items are specified early to meet the project schedule. However, the location of the equipment has not been determined. As a result, margins are factored into the design to ensure that the completed pump system is capable of meeting its process requirements.

Next, the total head required must be determined. This is done by arriving at the static and dynamic head values for the process elements.

Calculate the Static Head

In the example, the exact location of the supply tank and reactor have not yet been decided. The designers have an approximate size of the tank and vessel based on the process requirements. Until the civil department is able to complete the design of the tank foundations, the exact elevations and location of the supply tank and pressure vessel are estimated. The bottom of the supply tank is estimated to be at an elevation of 100 feet above sea level, but the bottom elevation of the tank could vary by plus or minus 10 feet. The estimated elevation of the nozzle into the reactor will be 180 feet above sea level, with a margin of plus or minus 10 feet.

Using the estimated elevation of the tanks, the elevation head for this system could be as low as 60 feet (170 feet – 110 feet) or as high as 100 feet (190 feet – 90 feet). The preliminary design will use 100 feet. Last, the operational pressure in the reactor is specified at 10 pounds per square inch gauge (psig) including the static head, but the operational pressure could vary by plus or minus 2.5 pounds per square inch (psi) from the design pressure. This pressure variation creates an additional 6 feet of pressure head. The resulting static head used in the preliminary design is 106 feet.

Determine the Dynamic Head

The next step is to determine the head loss in the pipeline as a function of the flow rate. This calculation incorporates the physical properties of the process fluid, the pipe material, the pipe length, and the valves and fittings.

Figure 1 includes all the isolation and check valves in the system. The design features these items to allow for the isolation of equipment for safe operation and maintenance. Other items, such as elbows and pipe lengths, are not indicated in the piping schematic. This exact information is not available until well after the preliminary design’s completion.

As a result, the pipe length and the number of fittings needed along with the design margin must be estimated. The pipe length between the pump discharge and process equipment is estimated at 150 feet, but a 50 foot margin is added. Further, a need for five elbows is estimated, with a margin of three more.

The final step in pipe sizing is calculating the head loss for each pipeline. The head loss is a function of the density and viscosity of the process fluid and the pipe inside diameter and roughness. The value for the pipe roughness and inside diameter is used in the head loss calculation. This information is referenced in pipe size tables for the selected pipe material.

When a pipeline is in service, it can corrode or foul over time. This results in an increase in pipe roughness and a reduction in the inside of the pipe diameter, leading to an increase in the pipeline’s head loss during operation. A design margin may be added for this unknown decrease as well.

The dynamic head for the system is calculated by adding the head losses of the individual pipelines to arrive at a total value of dynamic head. A design margin is often added.

The differential pressure for the process components must be factored into the pump sizing calculation. During the preliminary design, the process equipment has not been selected, so an estimated pressure drop must be used. The EPC has specification documents for the process equipment stating a maximum acceptable differential pressure in this system. For this equipment, the EPC allows a maximum of 15 psi. This estimate is converted to head and added to the dynamic head calculation for the process elements.

The sum of the static and dynamic head values is the head required by the process elements. Often, additional design margins are added based on experience. For example, pipe roughness may increase because of corrosion, or the pipe inside diameter may decrease because of sedimentation. This adds to the head loss in a pipeline but is hard to quantify, so a design margin may be added to take into account these unknowns.

The person performing the sizing calculations included these unknowns, but it was not documented. Because of this, another person added his/her own design margin for expected pipe corrosion. Additional design margins may be applied to the head loss during the review by the engineering manager.

The last item to be sized is the control valve. The mechanical group may add a typical differential pressure value of 15 psi, which is converted to head and added to the total head value for the pump.

With the results of the preliminary design complete, the EPC can develop a specification for the pump. The team specifies a flow of 800 gpm and a head of 211 feet of fluid.

Select the Pump

Next, the specification document is sent to the supplier for pump selection. Sales engineers often recommend pumps that have a best efficiency point (BEP) to the right of the design flow rate because they want to make sure that the selected pump can continue to meet the end user’s needs as the system capacity increases over time.

Figure 2. The pump curve selected for the application

Another common practice is to increase the impeller’s diameter to take advantage of the available motor horsepower. Electric motors come in different frame sizes based on the pumps’ power requirements. For example, the sample system’s pump has an impeller diameter of 14.75 inches, which will meet the end user’s need of 211 feet. However, if the impeller diameter is increased to 15.25 inches, the pump head can be increased to 228 feet of fluid without overloading the motor. The pump supplier can provide the customer with extra head to meet future needs without extra capital cost, but these practices increase the pump’s operating cost. Multiply this operating cost by the extra energy consumed, and the costs are significantly increased over the system’s life cycle.

This practice multiplied by the number of pumps in the plant can have a huge effect on the plant’s profit.

Examine the System Curve

After the preliminary design is completed and the equipment requirements are determined and recommended, the EPC team can see how the entire system works together (see Figure 3).

Figure 3. The pump system curve shows the interaction between the pump and the system, including the plant’s changing loads through the years.

During the first year of operation, the system is required to operate at 200 gpm. At the end of the second year, the system operates at 400 gpm. In five years, the plant is projected to operate at 800 gpm. The pump performance is displayed on the pump curve, shown in black. When the process operates at 200 gpm, the pump produces approximately 240 feet of head.

The system requirements are displayed in blue. The static head is shown where the blue line intersects the zero flow line. As the flow rate increases, the dynamic losses because of pipe friction and losses across the process equipment increase. At 200 gpm the system requires approximately 125 feet of head. The head loss across the control valve is the difference between the pump head and the system requirements. At 200 gpm the head loss across the control valves is about 115 feet of head.

The pump’s BEP occurs at 1,088 gpm, but the pump will be running at less than 400 gpm for the first few years of operation. After five years, the pump will still be operating at less than 80 percent of the BEP flow.

Conclusion

These are common practices used when sizing and selecting centrifugal pumps. Unfortunately, these practices are often repeated in the design phase and when the pumps are selected for process changes. The addition of design margins is an established and recommended practice. One way to improve the process is to document each time a design margin is used to ensure that it is not repeated by other engineering team members involved in the project. Design margins are intended to protect end users’ investments against unknown factors, but these practices increase the pump’s operating cost.

Next month’s column will examine this system after it has been placed in service and investigate what each of the added design margins actually costs. The column will also explore ways to incorporate design margins while minimizing increases to the operating, maintenance and capital costs.

Notes
Often people refer to an additional margin for unknown conditions as safety margins. These margins do not improve the safety of the system. As a result, the author uses the term design margin to describe the process of adding margins for unknown conditions.