1. 1 Comprehensive Pump Testing Challenges

2. 2 Purpose This presentation will discuss:  Hardships encountered while implementing comprehensive pump test requirements.  Benefits that may be realized if the comprehensive pump test (CPT) requirement is eliminated in favor of a Group A tests conducted at the same flow rate.  A historical review related to the origin of the CPT.

3. 3 Background The requirement to conduct a Comprehensive Pump Test became mandatory with the NRC endorsement of the 1995 Edition of the ASME OM Code, including the 1996 addenda. Since that time, the NRC has received many requests for relief, however, not all of these requests were for the anticipated reason.  The expected request dealt with the impracticality of performing the test due to the inability to achieve the desired flow rate.  The unexpected requests revealed that utilities did not see a corresponding increase in the level of safety or quality based on the additional costs associated with the development, implementation and maintenance of a new test and procurement of new instruments. This was based on the fact that most quarterly Group A tests were already being conducted at the required CPT flow rate.

4. 4 Background  The various issues associated with the CPT were brought to the attention of the ASME OM Code Committee.  The Committee agreed that changes were necessary.  Revisions to the Code have been proposed, but not approved.  A White Paper was developed to support the requested change.

5. 5 Background What are the issues? First, an attempt to understand why the CPT was developed is necessary.  It must be understood that this is my opinion only and is based on previous writings and attendance at OM Code meetings and symposiums since 1996.  I was not involved with ASME during the time that the original discussions began during the late 1980s and early 1990s.

6. 6 Background Papers were presented at the second and third NRC/ASME symposiums, and a letter was written by the NRC to the O&M Chairman regarding the need for a design basis test. These documents cover the late 80s and early 90s. The NRC wanted a test that was more effective in detecting pump degradation and that was capable of assuring the design basis capability of the pump.  Essentially, there was concern about the ability to correlate degradation at minimum flows to the operability.  The NRC was also concerned about pump damage that could occur by testing on minimum flow.  Both the NRC and the ASME Working Group felt that this new test needed to be practical and not task plant resources.  Because the test was more difficult, it would be conducted at an infrequent interval. It was decided that there was an urgent need to develop a comprehensive pump test.  This test would provide a better evaluation of pump characteristics at a reduced frequency. The OM-6 working group agreed to develop a CPT meeting the stated objectives. With this understanding, the NRC accepted the expanded upper hydraulic range from 3% to 10%. The expanded limits were introduced in OM-6.

7. 7 CPT Challenges As the CPT was being developed, ways to improve the test were reviewed and it was discovered that the only practical instrumentation improvements that could be made was associated with pressure (differential pressure).  Requirement was reduced from 2% to 1/2%. The hydraulic requirement was changed to require that the pump be operated at or near design flow.  Due to the hardship involved with installation of more accurate test instruments, the test would be performed during shutdown conditions.

8. 8 CPT Challenges Positives and Negatives associated with the CPT. Positives  Testing at ‘Design Flow’ makes it easier to determine the operability of a pump when degradation occurs.  More accurate pressure instrument results in more margin. A more accurate pressure gage reduces the amount of error included in development of the minimum performance requirement.  Typically, the accident performance requirement is artificially elevated to incorporate the error associated with the test instruments.

9. 9 CPT Negatives  Cost Plant modifications or processing of relief requests will be needed if the required test flow cannot be achieved.  Potential reduction in safety system availability Due to small hydraulic improvements.  Potential to increase dose and maintenance costs Due to small hydraulic improvements.  Increased procurement, implementation and maintenance costs More accurate pressure devices cost more. Additional procedures are necessary. More engineering work will be necessary.

10. 10 CPT Negatives  Potential for testing with less repeatability High accuracy pressure gages cannot be left in place and must be removed following testing.  Post maintenance test issues Related to establishing reference values for the associated Group A test. Which test do I run following maintenance?  More subjectivity What is the Engineer supposed to do in the event that a pump has passed the normally scheduled Group A test, but would have failed the CPT if it had been performed?

11. 11 CPT Challenges It is also important to point out that the pressure instrument is the ONLY equipment change This concept is commonly misunderstood.  Some documents indicate that the CPT requires the use of more accurate flow instrumentation (ref: NUREG/CP-0152, Vol. 4, page 3A-43).  This position has also been misstated in public meetings and forums. In reality, the flow rate, speed and vibration instrument requirements for a CPT and a Group A test are identical.

12. 12 CPT Challenges CPT improvements were limited to:  Requirement to test within 20% of design flow.  Pressure (differential pressure) gage accuracy improved to 0.5%.  The hydraulic required action limits were reduced from 1.10 x reference to 1.03 x reference. Note that this change only involves the upper required action range. The lower required action range for all pump types did not change.

13. 13 CPT Challenges There is no argument that testing a pump at substantial flow provides a better overall assessment of both the hydraulic and mechanical pump condition. With that understanding, many plants routinely test their pumps at substantial flow conditions consistent with the intent of a CPT, when practical. Therefore, if the substantial flow condition is being met, what advantage is there to conducting a biennial CPT?  The only outstanding differences are the accuracy of pressure gage and the difference in the upper acceptance limits.

14. 14 CPT Challenges Another change that is overstated deals with the use of the term ‘more restrictive acceptance criteria.’  The fact is that the limits for a CPT are not more limiting when compared to the Group A test with respect to degradation.  The lower bounding (required action) limits for a Group A test and a CPT are IDENTICAL.

15. 15 CPT Challenges The only hydraulic limit change involves a reduction from 10% to 3% of reference. The use of a 3% upper limit may impose a unwarranted problem for which there is no clear answer. With the restricted upper limit, it is quite possible to easily pass the routine Group A test, yet find that a CPT would have been unacceptable due to high flow rate or differential pressure. This places the engineer in a precarious position.  Initial guidance regarding this issue suggested that engineering judgment would apply.  More recent guidance provided at the 2004 NRC/ASME Symposium indicate indicates that the ‘problem’ should be resolved prior to the conduct of the CPT.

16. 16 CPT Challenges The fact is that there may be no ‘problem’ to correct. A 3% test deviation (or 1.03 x reference) is not unrealistic given the allowable ranges and instrument errors associated with an inservice test. As previously indicated, the Code instrumentation requirement for flow rate did not change.  The allowable flow error is 2% of full scale, with the full scale being limited to three times the reference value. Under worst case conditions, this could yield a worst case error of 6%.

17. 17 CPT Challenges A 6% flow rate error can easily cause a test to exceed the 3% upper limit. Bear in mind that other factors have not been considered in this scenario, such as:  Pressure indicator error and the effect of the performance point on the pump curve  Temperature drift  Mechanical error contributions (orifice plate, venturi tolerances, etc.)  Parallax error  M&TE calibration tolerances  Test data fluctuations  Meter readability – IEEE requirements of ½ the smallest increment  Allowable variance around the reference point.

18. 18 CPT Challenges In addition, if a pump fails a CPT due to exceeding the 3% upper limit, the pump must be declared inoperable and corrective actions implemented or the condition evaluated.  The net effect of this action is a reduction in safety!  Corrective actions will increase maintenance and operating costs and may introduce unwanted errors. Declaring a pump Inoperable as a result of an ‘improved’ hydraulic condition is a very difficult requirement to implement.  Given the fact that many IST pumps may DEGRADE at least 7% before actions are required, and centrifugal pumps can degrade by as much as 10%.

19. 19 CPT Challenges – The 3% upper limit There is no rational explanation as to why the Code would require a more critical assessment of pump performance due to elevated flow or pressure.  Excessive pump improvement can be indicative of a potential problem; however, these cases are extremely rare.  Trending of pump performance (increasing or decreasing) is now a requirement of the Code. The 3% upper limit was in place up until the adaptation of the 1988 ASME OM Code edition, at which point the upper hydraulic limit was increased from 3% to 10%. To supplement this change, a white paper was developed and discussed the changes and the basis for these changes.

20. 20 CPT Challenges – The 3% upper limit With regard to the elimination of the 3% upper hydraulic limit, the discussion centered on the fact that more emphasis was being placed on vibration measurement as the primary indicator of pump degradation. This was based on measuring vibration in velocity mode (inches/sec), vice displacement (mils). This change was brought about because there was concern relative to the ability to detect a change in pump performance based on hydraulic parameters. Consequently, it was determined that vibration measurement (using the new method) would be more sensitive to changes in pump performance.  It was stated that use of this technology would reduce the number of pumps requiring increased testing or corrective action based on erroneous (hydraulic) test results.

21. 21 CPT Challenges – The 3% upper limit Thus, the extent of the change was to allow equipment to be run in a “window” hydraulically and then to evaluate pump performance more closely with vibration. The window serves two purposes.  First, it ensures that the pump is performing its primary function of pumping liquid and  Operated in a narrow band where the vibration data will be repeatable. It was also recognized that positive displacement and vertical line shaft pumps could not be treated the same as centrifugal pumps. Thus, the hydraulic limits for degradation were not changed for these pump types.

22. 22 CPT Challenges – The 3% upper limit The 3% upper limit was re-instated in the 95 edition of the OM Code.  It has been published that the reduced upper limit ensures that the test results are not impacted by erroneous instrumentation.

23. 23 CPT Challenges The goal with the development of a CPT was to establish a more thorough and vigorous biennial test supplemented with less rigorous quarterly tests. However, if a Group A test is performed at the CPT flow rate, only the Group B pump test, by eliminating the requirement to measure vibration, would qualify as a less rigorous test. The Group B pump population is small.  Pumps that are operated only to implement IST requirements for periodic testing.

24. 24 CPT Challenges It is commonly stated that the CPT is a more rigorous test that is supplemented with a periodic Group A or B test using more relaxed acceptance limits and less rigorous requirements. In reality; The Group A and B test hydraulic acceptance limits for Operability DID NOT change from the previous edition of the Code. The Group A test mechanical acceptance limits for Operability DID NOT change from the previous Code edition.

25. 25 CPT Challenges The only pumps subjected to less rigorous testing were the Group B pumps by eliminating the requirement to measure vibration at a quarterly interval. The instrument accuracy requirements for flow rate, speed and vibration DID NOT change. All pumps were now required to be tested at design flow by development of the CPT.  This alleviates previous concerns related to correlation of minimum flow test results to design basis capability. Specifically, the effect of various degradation mechanisms on the shape of the curve could not be predicted.

26. 26 CPT Challenges – Pressure indication The pressure (differential pressure) instrument accuracy was reduced from 2% to ½%. The basis for this change was to provide more accurate reference values.  It was published that the instrumentation requirements were identical to those established for the preservice test.  As a result, it would be easier to detect degradation during subsequent tests based on the fact that you would be comparing pump performance using the same instrumentation and reference point that was established during the preservice test.  It has been stated that another driving force behind the ½% gage was due to the fact that the test was only conducted every two years. Does the improved pressure instrument accuracy provide better long term trending capability? In my opinion……… NO!

27. 27 CPT Challenges – Pressure indication Here’s why? Repeatability – Trend capability Sensitivity Actual accuracy A more accurate pressure instrument may yield less positive attributes with respect to repeatability.  The gages are more sensitive and prone to failure if left in service; therefore, are removed and transported to and from the test site.  The gage is then installed, vented and zeroed (for many test gages) prior to use. Interpretation of the result may also lead to inconsistencies as a result of the increased sensitivity of the device, which would tend to result in more active needle oscillations, forcing the reader to employ averaging techniques if the oscillations cannot be dampened.  The increased sensitivity also increases the risk of impacting the calibration of the instrument due to physical agitation when in transit, installed or removed.

28. 28 CPT Challenges – Pressure indication This could result in a calibrated gage appearing uncalibrated, or vice versa. In short, a test using a more rugged, permanently installed gage may offer better long term trend capability when compared to the more accurate, but less rugged test gage. Realized accuracy  Although installed gages are certified to an accuracy of 2%, I have found that they are actually accurate to at least ½% based on a review of as-found calibrations and discussion with our I&C department.  In fact, we are looking at increasing our calibration interval based on the success of our as-found calibration checks.  It would be interesting to see if a review of industry calibration records yield a similar result.

29. 29 CPT Challenges - Summary Implementation of the CPT requirements may produce unwanted outcomes for those utilities that already meet the primary intent of a CPT by testing their pumps at design flow. When you update your program and begin implementing the CPT requirements, you can expect the following:  Challenges from the plant staff (Operations, Maintenance, Procedures, Management) regarding the value of this additional test.  Additional workload associated with maintaining and updating two sets of reference values and assignment of post maintenance test requirements.  Emergent work and reduced safety system availability – At RNP, the upper acceptance limit has been exceeded on two occasions. This required that pumps be taken out of service. See bullet 1 for additional effects.  Budget for additional cost – RNP spent approximately $22,000 on new pressure gages.  NRC denial of your proposed request to use the Group A test at the CPT flow rate in lieu of the CPT.

30. 30 CPT Challenges - Summary It is my belief that a quarterly Group A test conducted at the CPT flow rate is more effective in evaluating pump performance and detecting degradation compared to the conduct of a CPT at a biennial interval supplemented with a quarterly test at a lower flow rate.  The pump is tested to an identical hydraulic load.  Vibration, flow rate and speed instrumentation requirements are identical.  The lower bounding hydraulic limits are identical.  The mechanical limits are identical.  More frequent testing at higher flow rates (using the same test equipment) provides a much better trend capability.

31. 31 CPT Challenges - Summary If a quarterly test is conducted at the same flow rate as a CPT, then: The primary concerns related to verification of operability and detection of degradation are no longer valid.  This was the primary motive for developing the CPT. A test gage is not necessary.  Since the equivalent test is being performed quarterly, there would be no need to obtain ‘more accurate’ data every two years.  In reality, the more accurate gage may not offer better accuracy, or repeatability.

32. 32 CPT Challenges - Summary The reduced upper hydraulic acceptance limit is not necessary.  Because the equivalent test is performed quarterly, there is no need to tighten the upper limit to counter the effects of instrument error.  There is no equivalent criteria for degradation, which is the prevalent failure mechanism.  You are required to trend performance.  The improved vibration requirements that resulted in the previous expansion of the upper limit are still in place.  The upper criteria could be exceeded due to normally expected instrument error allowances.

33. 33 CPT Challenges - Benefits The benefits associated with this proposal are many. First and foremost:  Better assessment of overall pump performance.  Better trend capability.  Improved ability to detect degradation.  Reduced capital costs. Accurate pressure gages are expensive.  Reduced O&M costs. Installation and maintenance of test gages. Develop, maintain and update additional procedures. Engineering evaluation using two sets of criteria.

34. 34 CPT Challenges - Benefits Focus available resources on more urgent matters.  The industry has been downsized.  Evaluations, maintenance, etc. should not be mandated based on small increases in performance based solely on test data, unless there is a valid concern supported by additional data (more tests, vibration data).

35. 35 CPT Challenges - Benefits Increased safety system availability.  If the upper limit is exceeded, the pump is inoperable.  It cannot be returned to service unless maintenance is conducted or the condition is evaluated. New reference values must be established if an evaluation is used to return a pump to an operable status. New reference values established at a higher value may not be the best long term solution, if the test result is not typical of normal performance and is not supported by additional data.

36. 36 CPT Challenges - Conclusion The Code should be used as a tool to:  Determine and assess overall pump health.  Trend performance and take actions prior to failure.  Ensure operational readiness and provide one of the key inputs necessary to determine operability. A quarterly test conducted at or near design flow accomplishes these objectives.  Also, a full flow test conducted 8 times over a two year period compared to one time in a two year period is far more effective in evaluating overall pump health, provides better trend capability, and provides more frequent assurance of meeting design requirements.

37. 37 CPT Challenges - Conclusion The Code should not:  Place unwarranted burden upon the utility. This was one of the stated objectives when developing the CPT. If your Group A tests are being conducted at design flow rates, then the additional comprehensive test may be considered a burden.  Bring acceptable test results into question. A Group A test that passes at 104% of the allowable limit will result in questions relative to operability.  When compared to the CPT limit. This tends to lead to passionate debates and is very difficult to defend. It is also difficult to tell management that the pump may have to undergo maintenance to ensure that it passes the next scheduled test.

38. 38 CPT Challenges - Conclusion QUESTIONS?