Luis API 571 & API 581 RBI guidelines is OK to start a survey for TML. Pipeline inspection should be based on the following selection criteria- from my experience in monitoring process chemical, fertilizer, refinery /petro-chemical plants. Your first critical selection should be based on: 1. Mmaterial used, 2. Service duty 3. Service temperature 4. Service pressure in the system, 5 Service velocity or flow in the piping system, 6. Service obstacle in the system like orifice, bends, elbows or Tees (in other words look into possibility of change of direction or flow,velocity increase that can affect flow), 7. Service change of direction - look at branch connection areas, 8. Service Corrosion Allowance given by piping design (likely expected corrosion or mechanical damage in the system from previous 9. Acutal or reported experience of other similar users, 10. Service flow or temperature control valves in the system and their effect in throttling flow or temperature effect on both upstream or downstream side of the piping system). After this, choose critical monitoring locations to get reasonable Corrosion mapping profile in your first survey viz., on each bend, elbows, straight pipe (upstream and downstream of pumps, control valves,orifice plates, branching lines including small lines) should be done on all directions to get a reasonable profile of the selected locations in your TML. After this first survey, calculate the discard thickness taking into account the corrosion allowance permitted as per piping design or piping code for each line to ASA B 31.3 or API for the particular service duty, Based on discard value and the profile pattern obtained in the first survey of selected line - arrive at the possible variations - after discounting the variations allowed for each pipe supplier or manufacturer as per piping code supply tolerances ( pipe can be supplied to 12.5% of variations in thicknessas per piping code). Once it is established that variations obtained in the first survey of random and pre-determined locatins show some distinct signs of corrosion (based on the above factors for first monitoring schedules), you would be reasonable sure that some locations need more careful monitoring for the next survey. In other words these locations would show distinct changes which could be outside the allowed pipe thickness tolerancesand also above the admissible C.A for this pipe. With this, your choice for critical monitoring for the second and third inspection schedules would become easier. Your next attention or priority should be to first re-confirm for the second or thrid survey that the monitored locations with variations are maintaining the same values and re-establish that you have not missed the same locations monitored in the earlier year. Once two or three successive inspection have established that the thickness variation observed is either within the C.A limits or outside the C.A. linits then apply a statistical profile minimum discard or minimum thickness pattern for the particular section for replacement and fix or recommend a time schedule. With this approach, you could come for a statistical remaining life prediction and apply API 581 RBI norms. You would get more or less accurate remaining life. API 581 or API 570 may give braod recommendatory guideline to follow. Your monitoring or in-plant experience or information from experience drawn from other similar users or other practising engineer guideline would be a valuable guideline or help a lot in this RBI prediction monitoring and more accurate prediction. Trust this is of help to you C.V.Srinivasan Nishi Engineers Pvt Ltd India Feb 23,2006 E-mail:nishi@vsnl.com > > >The selection of our thickness measurement locations (TMLs) >is established in accordance with the API 570 guidelines. > >Critical Piping circuits are selected in accordance with >RECOMMENDED PRACTICE 571 Damage Mechanisms Affecting Fixed >Equipment in the Refining Industry and API RECOMMENDED >PRACTICE 580 Risk-based Inspection. > >After the base line survey and the first inspection we >calculate a corrosion rate based on the TML minimum >thickness found in the circuit. With this corrosion rate we >define the next inspection survey to define inspection >frequency and estimation of the remaining life. After de >third inspection survey we are in conditions to define a >short term and a long-term corrosion rate, inspection >frequency and a more accurate remaining life. > >My question is: > >Are there any guidelines for the statistics analysis of TML >UT measurements? > >A piping circuit section should be replaced when a TML of >that section reaches the minimum calculated thickness or it >should be given a statistics analysis of all TMLs within the >whole critical piping circuit to preview the corrosion trend >and pipe length replacement? > >Comments or guidelines on this subject should appreciated > >Regards > >Luis Marques