September 05, 2007

How to measure pump performance without flowmeters?

We all face a very common problem of how to measure the performance of an inefficient pump in absence of flow meters particularly in case of cooling water pumps where usually there is no flow meter.

Based on my past experience in energy audit for last several years I found that we know the pump is not upto the mark based on its physical apperance of corrosion, leakages, sound etc. but how to prove it based on data that too when it is delivering the design head when you measure discharge & suction pressures. This is very confusing.

Read the simple method to verify the performance. You only need the performance curve originally supplied by the manufacturer.

This is a very common problem that all of process engineers face. How to estimate the performance sufficient enough to make a decision on improvement measures be it overhauling, replacement etc.

The confusing thing is that we are able to see the problems physically, the corrosion, leakages, sound, extra power consumption BUT still feel doubtful as pressure gauge is showing a good reading near to design value on the discharge side.

If we consider this discharge head & plot it on curve to find out the flow rate it is also similar or near to design value giving a good efficiency on mathematics. This is the point where most of us fails. So what to do?

My simple approach is to start measuring from shut off head.

  1. First Read the curve for Q & H and put it in Excel table.

  2. Use simple X-Y plot to draw this H Vs Q curve.

  3. Try to fit this data in a cubical equation, which is sufficient mostly with R-sqaured = 0.999 or better generally.

  4. Calculate H value for each Q in the same table based on curve fit equation & check for any significant deviation.

  5. If deviation is found fine tune your fit data either by dividing flow rate or by multiplying using a common factor which should be incorporated in the final equation. This is a problem related to Excel as it gives large numbers in exponential form after truncation. Otherwise, if you are using any other data fit software you don't need to do it.

  6. Now start your experiment.

  7. Measure shut off head & power at this condition.

  8. Measure discharge pressure at several flow rate values by opening the discharge valve slowly. Also record the power for each point. You should consider ~4-5 points including 100% full open condition.

  9. At least two points are absolutely necessary with shut off & 100% opening.

  10. Plot these points along with original design data.

  11. Calculate flow rate from the curve based on equation derived above but considering shift in shut off head.

  12. This flow rate value is the MAXIMUM possible value of flow which your pump can give in the given operating condition.

  13. Actual flow rate will be further less depending on the corrosion, leakages etc.

  14. However, you can judge the performance which may be sufficient enough to decide the replacement or modification in the system.

  15. This will also give you an idea about your system requirement - static head and dynamic head both. Thus you can also decide if a pump with lower head is suitable or not.

Print this post


Rhen Nicey said...

Whether you have working in industry or you are in a place where you need to look into finding real solutions for real problems, you may find that flowmeters are one tool that you should consider. Thanks for this post :)

Total Pageviews

Support Us

If you find this Blog useful Kindly take your time to donate some amount to keep it running.