Page 92 - Subyek Teknik Mesin - Forsthoffers Best Practice Handbook for Rotating Machinery by William E Forsthoffer

Be st Practice 2 .12 Pump Best Practices
Field operation
Step Action
1. Confirm NPSH A NPSH R at maximum operating flow. If margin The preceding section described requirements to ensure opti-
less than two (2) feet, require witnessed NPSH R test. If NPSH R mum pump reliability during the project design phase. This
> NPSH : section will describe how to detect and correct hydraulic dis-
A
Increase NPSH A by: turbances in the field.
Increasing suction drum level
Decreasing pumping temperature
Decreasing suction line losses Determining the potential for damage
Reselect pump (if possible)
Select canned pump
2. For the pump selected calculate N SS based on pump BEP Hydraulic disturbances are detected by monitoring the condi-
conditions. Note: if double suction first stage impeller, use 1/2 tions noted in Figure 2.12.26.
of BEP flow
3. If N SS > 8000, contact pump vendor and require following data
for actual pump fluid and conditions:
Loud noise – continuous or varying
Predicted onset flow of cavitation caused by recirculation for
Suction and discharge pressure pulsations
actual fluid conditions
Reference list of proposed impeller (field experience) High values of overall vibration
4. Compare cavitation flow to minimum operating flow. If this Drop in produced head of > 3%
value is within 10% of minimum operating flow: High values of vane passing frequency in overall spectrum
Reselect pump if possible Pressure pulsations in inlet and discharge piping
Install minimum flow bypass Possible high bearing temperatures
Consider parallel pump operation
5. Calculate liquid temperature rise at minimum operating flow. If
value is greater than 5% of pumping temperature: Fig 2.12.26 Indicators of hydraulic disturbances
Calculate NPSH A based on vapor pressure at calculated
pumping temperature. If NPSH A < NPSH R : Once detected, the root cause of the disturbance must be
Install minimum flow bypass established. As previously discussed, vaporization of liquid
Consider parallel pump operation
must be present for cavitation to occur, so the requirement is
to determine the root cause of vaporization. As previously
Fig 2.12.24 Guidelines for selecting pumps free of hydraulic discussed, there are three (3) primary root causes of
disturbances vaporization:
Internal inlet pressure losses
Formation of low pressure cells at low flows
‘Since the actual liquid characteristics are not known when
the standard pump curve (tested on water) is drawn, the vendor Liquid temperature rise at low flows
stops the NPSH R curve where flow separation and liquid Confirmation of process conditions and specific tests are re-
temperature rise can cause liquid disturbances’. quired to determine the root cause of the problem.
Therefore, trouble free operation to the left of this point is
dependent on the pumped liquid and must be discussed with
the pump vendor. Determining the cause of hydraulic
In conclusion, preventing liquid disturbances in the project disturbances
design phase requires a thorough, accurate investigation of both
the process and pump characteristics and some serious decisions Confirmation of stated value of NPSH
on required action. A
Justification of required action will be easier if the operating
company looks beyond the project costs and examines what the Refer to the Pump Data Sheet and/or the Hydraulic Calculation
Sheet for that pump service to determine the stated NPSH A .
actual cost (lost in production) will be if a pump experiences Proceed to check the NPSH A at field operating conditions.
hydraulic disturbances. Most operating companies have docu- Using the relationship:
mented case histories of problem pumps that will provide
proven facts relating to the ‘total cost’ of operating problem
2:311 ðP suction PSIA P vapor pressure PSIA Þ
pumps for the life of a project (refer to Figure 2.12.25). NPSH A ¼
S:G: at pumping temperature
Substituting in the equation the following actual values:

Determine the cost effectiveness of pump selection not Pump suction pressure (down stream of suction strainer)
only on the project (capital investment) costs, but on the Actual vapor pressure at measured pumping temperature
cost to the operating company of unreliable pumps Actual S.G. at measured pumping temperature
Compare calculated NPSH A to predicted NPSH A . If actual
Fig 2.12.25 Justifying the selection of troublefree pumps NPSH A < predicted NPSH A , modify operation if possible to

66