jbuchyns
Civil/Environmental
- Sep 17, 2001
- 4
The Question is:
Has anyone conducted or know of a study on the recommended frequency of performance tests, especially for large water pumping systems, as a means of predicting maintenance requirements?
--------------------------------------------------------
My group is responsible for conducting performance testing on large centrifugal pumps for a major water utility. We are currently establishing a baseline of performance and efficiency curves for all (100+) of our units, in order to monitor performance degradation over time. The information from this test database will be used for maintenance scheduling and energy management, including optimization.
One thing we need to establish is the required frequency of this kind of testing. We know that these units degrade over time, and we are treating performance testing as a maintenance evaluation tool prior to calling for any kind of overhaul. The expectation is that we will be able to do a trend analysis on the degradation in order to predict the optimum timing of any maintenance event versus the increasing operating costs over time.
We're starting with a test trigger of 5,000 hours runtime, which will certainly keep my team busy. This may be a little too frequent, but it's a starting point for us and we figure the first dozen or so re-tests should give us an idea of how to adjust this trigger.
Online hydraulic performance monitoring via a new process control system will allow us to identify units which are slipping by calculating wire-to-water efficiencies. This will help to readjust the test frequency as well as providing us with an additional triggering event. As well, we will be able to monitor bearing Delta Tº online.
We are also using vibration monitoring and oil analysis strictly as maintenance tools, and want to work in performance testing as an additional resource.
Has anyone conducted or know of a study on the recommended frequency of performance tests, especially for large water pumping systems, as a means of predicting maintenance requirements?
--------------------------------------------------------
My group is responsible for conducting performance testing on large centrifugal pumps for a major water utility. We are currently establishing a baseline of performance and efficiency curves for all (100+) of our units, in order to monitor performance degradation over time. The information from this test database will be used for maintenance scheduling and energy management, including optimization.
One thing we need to establish is the required frequency of this kind of testing. We know that these units degrade over time, and we are treating performance testing as a maintenance evaluation tool prior to calling for any kind of overhaul. The expectation is that we will be able to do a trend analysis on the degradation in order to predict the optimum timing of any maintenance event versus the increasing operating costs over time.
We're starting with a test trigger of 5,000 hours runtime, which will certainly keep my team busy. This may be a little too frequent, but it's a starting point for us and we figure the first dozen or so re-tests should give us an idea of how to adjust this trigger.
Online hydraulic performance monitoring via a new process control system will allow us to identify units which are slipping by calculating wire-to-water efficiencies. This will help to readjust the test frequency as well as providing us with an additional triggering event. As well, we will be able to monitor bearing Delta Tº online.
We are also using vibration monitoring and oil analysis strictly as maintenance tools, and want to work in performance testing as an additional resource.
