Predict efficiency using the Hydraulic Institute’s standards as guidelines.

Energy conscious centrifugal pump users are always interested in knowing the value of maximum attainable pump efficiency for the required capacity and head. This value can be used as a benchmark to facilitate:

  • Selecting and purchasing an energy efficient pump 
  • Conducting an energy audit of pumps in an operating plant/refinery to make decisions about the replacement of energy inefficient pumps 

In view of these advantages, developing a simple method to predict maximum attainable efficiency was needed. 

Historical Performance Curves

Approaches such as referring to pump hydraulic performance curves available in the archives of a refinery/plant or on manufacturers’ websites have limitations. These limitations include the availability of limited data and the retrieval of efficiency values from this data. This retrieval is mostly a manual process, which is time consuming and cumbersome, especially when the number of pumps is large. 

Further, the value retrieved from this limited data may not be a true representation of the maximum attainable efficiency. The possibility of using a single equation or charts available in literature [1, 2, 3, 4, 5 & 6] was checked. The search for a single, reliable equation covering a wide capacity range and pump types proved futile. The comparison of charts available in the literature [2, 3, 4, 5 & 6] for efficiency prediction revealed that those given in Figure 1A and 1B of Efficiency Prediction Method for Centrifugal Pumps [6], offered several advantages and were selected. 

Hydraulic Institute’s Method

In the seven graphs in Figure 1A from Reference 6, the efficiency (at optimum specific speed) is expressed as a function of capacity. The Hydraulic Institute’s (HI) presentation makes efficiency reading easy because pump users’ are accustomed to reading efficiency as function of capacity, which is always available. In other references [3, 4 & 5], the efficiency, at constant capacity, is expressed as a function of specific speed. This makes efficiency reading difficult because most of the time specific speed values require the operating speed [7] for calculations. This is usually not readily available to the end user. The seven graphs in Figure 1A cover a wide range of pump designs. Each of the seven graphs refers to a particular design. 

This segregation based on design coupled with the specific speed correction chart in Figure 1B helps the pump user generate more options for a particular requirement of head and capacity. More options help the user make informed choices as explained in “The Energy Audit” section. The charts in Reference 2 also had potential but were not considered because of the limited application range. 

Excel Spreadsheet to Speed Calculations

Based on the charts from Figure 1A, an interactive Excel spreadsheet was developed to help make the efficiency calculations faster. To check the reliability of the charts, the efficiency values calculated using them were compared with the pumps available in the market. This comparison, shown in Table 1, was carried out for most of the widely used pumps: API overhung center-line supported (OH2 Type) pumps designed as per API [7] in capacities ranging from 18 cubic meters per hour (m3/h) to 1,080 m3/h. The figures in Columns D and E in Table 1 reveal that reliability of the prediction is higher if the pumps’ specific speeds are closer to 2,500 rpm. 

Higher efficiencies are observed when the specific speed is near 2,500 rpm. This suggests that pump users should try to purchase pumps that have specific speeds close to the optimum specific speed of 2,500 rpm. Pump users can achieve this to some extent by selecting the right combination of speed, head/stage and type as indicated in Table 2. However, in actual applications, selecting a pump using these criteria is not always practical. In such cases, the pump user can use HI’s predicted efficiency values at optimum specific speed, as well as with specific speed corrections. This will give a reasonable understanding of efficiency for pump selection. These values provide a possible range of efficiency and power savings for making an informed decision. 

Please feel free to email the authors if you would like a spreadsheet for review.

The Energy Audit

Considering the limitations, an energy audit was conducted on the pumps in an operating refinery. Some of the audited pumps’ data are shown in Table 2. The efficiencies of the pumps (those predicted with the HI method for different design options and those per other approved vendors are mentioned. The possible energy savings are also noted. The power consumption details will facilitate life-cycle cost studies to decide about the replacement or upgrading of existing pumps. 

The column “Comments” provides possible actions. The interacting spreadsheet can be used to generate options (Table 2) to make more informed choices. These options (for example, Case 1) show the possibility of using cheaper, overhung pumps with almost the same efficiency (60 percent) or between bearing pumps with more stages with considerably higher efficiency (68 percent) for existing between bearing pumps. The market availability of pumps with efficiencies close to or better than what is predicted by spreadsheet proves the practical application of the spreadsheet. Most important, the spreadsheet indicates the possibility of using single-stage overhung pumps to replace costly between bearing pumps, which is evident from Case 6.

The spreadsheet, however, has certain limitations:

    • It is observed that, in some cases, vendors can offer better efficiencies than projected with this spreadsheet.
    • The flow range covered by HI is less than what is available in the market. The API overhung type pumps available in the market cover flow ranges up to 2,500 m3/h while HI covers those up to 1,100 m3/h.
    • Some factors—such as temperature and viscosity corrections, wear ring clearance, surface finish and suction specific speed discussed in Reference 6—are not considered. It is impractical for users to consider all these issues. 
  • Automatic generation of all possible options using different speeds, stages and types for a given capacity and head is not possible. A computer program in higher-level language, such as Visual Basic, is required.
  • It was possible to check the reliability for overhung and between bearing pumps only because these are the most widely used pumps in refineries. Checking other types of pumps (slurry, paper stock and sewage pumps) was not carried out because little data was available.

 

For a copy of the spreadsheet referenced in this article, email one of the authors.

Pumps & Systems, April 2012

Reference:

  1. Alejandro Anaya Durand; “A quick estimate for centrifugal – pump efficiency”; Chemical Engineering, July 1989.
  2. European Guide to Pump Efficiency for Single Stage Centrifugal Pumps; May 2003.
  3. Eugene P Sabini & Warren H. Fraser; “The effect of specific speed on the efficiency of single stage centrifugal pumps”; Indian Pumps, December 1986.
  4. Sulzer Centrifugal Pump Handbook; ISBN 1-85166-442-4.
  5. Pump Handbook, Ed 4, McGraw Hill Publication; ISBN No 978-0-07-146044-6.
  6. Efficiency Prediction Method for Centrifugal Pumps; Hydraulic Institute 1994.
  7. ANSI / API STANDARD 610: Centrifugal Pumps for Petroleum, Petrochemical and Natural Gas Industries, 11th Ed.

Acknowledgements:

The authors would like to thank the management of Essar Oil Limited and Dr. Joe Evans ( www.PumpEd101.com) for all their support and encouragement for this publication. They would also like to thank their colleagues Ashita Vaish, Priya Sharma, Shyam Bhat, Romal Chafle and Rahul Shende for their help in the data analysis.