Risk Based Inspection of Gas-Cooling Heat Exchanger

 PHE – ONWJ platform personnel found 93 leaking tubes locations in the fin fan coolers/ gas-cooling heat exchanger. After analysis had been performed, the crack in the tube strongly indicate that stress corrosion cracking was occurred by chloride. Chloride stress corrosion cracking (CLSCC) is the cracking occurred by the combined influence of tensile stress and a corrosive environment. CLSCC is the one of the most common reasons why austenitic stainless steel pipework or tube and vessels deteriorate in the chemical processing, petrochemical and maritime industries. In this research purpose to determine the appropriate inspection planning for two main items (tubes and header box) in the gas-cooling heat exchanger using risk based inspection (RBI) method. The result, inspection of the tubes must be performed on July 6, 2024 and for the header box inspection must be performed on July 6, 2025. In the end, RBI method can be applicated to gas-cooling heat exchanger. Because, risk on the tubes can be reduced from 4.537 m 2 /year to 0.453 m 2 /year. And inspection planning for header box can be reduced from 4.528 m 2 /year to 0.563 m 2 /year.

Based on the explanation above, Pertamina PHE-ONWJ gas cooling heat exchanger classified as areal cooler heat exchanger because its function is cooling the gas with a fan in to near ambient temperature.
Heat exchanger is the one of crucial equipment in the processing facility especially in the oil and gas industry sector. Heat exchanger is used to transfer heat between one and more fluids. Ones of heat exchanger application is for cooling the gas before injected to the oil reservoir. Gas injection is the method to increase oil production by boosting depleted pressure in the reservoir (figure 2). Another function of gas cooling heat exchanger is for cooling the gas before supply the gas turbine to generated electric power on the platform American Petroleum Institute (API) is the one of the most widely used standard guideline in oil and gas company around the world besides DNV-GL. PHE ONWJ platform adopt guidelines from API 660 and API 661 for gas cooling heat exchanger fabrication and installation. One of maintenance strategies for gas cooling heat exchanger can be developed by using Risk Based Inspection (RBI). by using RBI company will get information using risk analysis to develop an effective inspection plan.
International Journal of Marine Engineering Innovation and Research, Vol. 1 (4), Sept. 2017. 317-329 (pISSN: 2541-5972, eISSN: 2548-1479 318 Identification of company equipment is the beginning of the systematic process in the inspection planning. Probability of failure and consequence of failure are the basic formula to calculate the RBI and must be evaluated by considering all damage mechanism directly effect to the equipment or the system. However, failure scenarios according to the actual damage mechanism should be develop and considered.
RBI methodology produces optimal inspection planning for the asset and make the priority from the lower risk to the higher risk. In other word inspection planning in RBI focused to identification what to inspect, how to inspect, where to inspect and how often to inspect. Inspection planning used to control degradation of the asset and the company will get considerable impact in the system operation and the appropriate economic consequences [2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18].

II. METHOD
The information of inspection planning in risk based inspection based on the risk analysis of the equipment. The purpose of the risk analysis is to identify the potential degradation mechanisms and threats to the integrity of the equipment and to assess the consequences and risk of failure [3].

A. Risk
Risk is defined as the combination probability of asset failure and consequence if the failure happened. Risk can be expressed numerically with formula (1) as shown below.

Probability of Failure
The probability of failure may be determined based on one, or a combination of the following methods: -Structural reliability models In this method, a limit state is defined based on a structural model that includes all relevant damage mechanisms, and uncertainties in the independent variables of this models are defined in terms of statistical distributions. The resulting model is solved directly for the probability of failure.
-Statistical models based on generic data In this method, generic data is obtained for the component and damage mechanism under evaluation and a statistical model is used to evaluate the probability of failure.
-Expert judgment In this method, expert solicitation is used to evaluate the component and damage mechanism, a probability of failure can typically only be assigned on a relative basis using this method.
In API RBI, a combination of the above is used to evaluate the probability of failure in terms of a generic failure frequency and damage factor. The probability of failure calculation is obtained from the equation (2).
Where: gff = generic failure frequency Df (t) = damage factor FMS = management system factor

B. Generic Failure Frequency (gff)
The generic failure frequency can be determined by asset failure of common industries. The generic failure frequency is expected to the previous failure International Journal of Marine Engineering Innovation and Research, Vol. 1 (4), Sept. 2017. 317-329 (pISSN: 2541-5972, eISSN: 2548-1479 319 frequency to any specific damage happening from exposure to the operating environment. There are four different damage hole sizes model the release scenarios covering a full range of events they are small, medium, large, and rupture. If the data of the asset is complete, actual probabilities of the failure could be calculated with actual observed failures. Even if a failure has not occurred in a component, the true probability of failure is likely to be greater than zero because the component may not have operated long enough to experience a failure. As a first step in estimating this non-zero probability, it is necessary to examine a larger set of data of similar components to find enough failures such that a reasonable estimate of a true probability of failure can be made.
This generic component set of data is used to produce a generic failure frequency for the component. The generic failure frequency of a component type is estimated using records from all plants within a company or from various plants within an industry, from literature sources, and commercial reliability data bases. Therefore, these generic values typically represent an industry in general and do not reflect the true failure frequencies for a specific component subject to a specific damage mechanism.
The generic failure frequency is intended to be the failure frequency representative of failures due to degradation from relatively benign service prior to accounting for any specific operating environment, and are provided for several discrete hole sizes for various types of processing equipment (i.e. process vessels, drums, towers, piping systems, tankage, etc.). A recommended list of generic failure frequencies is provided in Table 1. The generic failure frequencies are assumed to follow a log-normal distribution, with error rates ranging from 3% to 10%. Median values are given in Table 1. The data presented in the Table 1 is based on the best available sources and the experience of the API RBI Sponsor Group.
The overall generic failure frequency for each component type was divided across the relevant hole sizes, i.e. the sum of the generic failure frequency for each hole size is equal to the total generic failure frequency for the component. C. Management System Factor Management system factor used to measure how good the facility management system that may arise due to an accident and labor force of the plant is trained to handle the asset. This evaluation consists of a series of interviews with plant management, operations, inspection, maintenance, engineering, training, and safety personnel.
The management systems evaluation procedure developed for API RBI covers all areas of a plant's PSM system that impact directly or indirectly on the mechanical integrity of process equipment. The management systems evaluation is based in large part on the requirements contained in API Recommended Practices and Inspection Codes. It also includes other proven techniques in effective safety management. A listing of the subjects covered in the management systems evaluation and the weight given to each subject is presented in Table 2.
The management systems evaluation covers a wide range of topics and, as a result, requires input from several different disciplines within the facility to answer all questions. Ideally, representatives from the following plant functions should be interviewed: The scale recommended for converting a management systems evaluation score to a management systems factor is based on the assumption that the "average" plant would score 50% (500 out of a possible score of 1000) on the management systems evaluation, and that a 100% score would equate to a one order-of magnitude reduction in total unit risk. Based on this ranking, equation (3) and equation (4) may be used to compute a management systems factor, , for any management systems evaluation score.   *Note that the management score must first be converted to a percentage (between 0 and 100) as follows:

D. Thinning Damage Factor
The calculation procedures of thinning damage factor are: a) Determine the number of inspections, and the corresponding inspection effectiveness category for all past inspections. Combine the inspections to the highest effectiveness performed. b) Determine the time in-service (age) since the last inspection thickness reading (t rd ). c) Determine the corrosion rate for the base metal (C r,bm ) based on the material of construction and process environment, where the component has cladding, a corrosion rate (C r,cm ) must also be obtained for the cladding. d) Determine the minimum required wall thickness ( per the original construction code or using API 579. If the component is a tank bottom, then in accordance with API 653 ( = 0.1 in) if the tank does not have a release prevention barrier and ( = 0.05 in) if the tank has a release prevention barrier. e) For clad components, calculate the time or age from the last inspection required to corrode away the clad material, , using equation (5).
f) Determine the parameter using Equation below, based on the age and from step b, from step c, from step d and the age required to corrode away the cladding, , if applicable from step e. For components without cladding, and for components where the cladding is corroded away at the time of the last inspection (i.e. = 0.0), use Equation (6).

E. Stress Corrosion Cracking Damage Factor
The calculation procedures of chloride stress corrosion cracking (CL-SCC) damage factor are: a) Determine the number of inspections, and the corresponding inspection effectiveness category for all past inspections. Combine the inspections to the highest effectiveness performed. b) Determine the time in-service (age) since the last Level A, B, C or D inspection was performed. c) Determine the susceptibility for cracking using Table 3 based on the operating temperature and concentration of the chloride ions. Note that a HIGH susceptibility should be used if cracking is known to be present.  f) Calculate the escalation in the damage factor based on the time in-service since the last inspection using the age from step b and equation below. In this equation, it is assumed that the probability for cracking will increase with time since the last inspection as a result of increased exposure to upset conditions and other non-normal conditions. = (age) 1.1 (8)

F. Consequence Analysis
The calculations of consequence procedures are: a) Select a representative fluid group from Table 6.
c) Determine the steady state phase of the fluid after release to the atmosphere, using Table   8 and the phase of the fluid stored in the equipment as determined in step b. d) Based on the component type and Table 9, determine the release hole size diameters (d n ). e) Determine the generic failure frequency (gff n ), and the total generic failure frequency from this table or from equation (10).
i) Group components and equipment items into inventory groups using Table 10.

Type of Detection System Detection Classification
Instrumentation designed specifically to detect material losses by changes in operating conditions (i.e., loss of pressure or flow) in the system A Suitably located detectors to determine when the material is present outside the pressure-containing envelope B Visual detection, cameras, or detectors with marginal coverage C

Type of Isolation System Isolation Classification
Isolation or shutdown systems activated directly from process instrumentation or detectors, with no operator intervention A Isolation or shutdown systems activated by operators in the control room or other suitable locations remote from the leak B Isolation dependent on manually-operated valves C   Table 11 select the appropriate classification (A, B, C) for the isolation system. u) Using Table 12 and the classifications determined in step s & t, determine the release reduction factor, fact di . v) Using Table 13 and the classifications determined in step s & t, determine the total leak durations for each of the selected release hole sizes, ld max,n . w) For each release hole size, calculate the adjusted release rate (rate n ) using equation (17) where the theoretical release rate (W n ).
x) For each release hole size, calculate the leak duration (ld n ) of the release using Equation 4.13, based on the available mass (mass avail,n ), and the adjusted release rate (rate n ) from step. Note that the leak duration cannot exceed the maximum duration (Id max,n ) determined in step w.
(18) y) For each release hole size, calculate the release mass (mass n ), using equation (19) based on the release rate (rate n ), the leak duration (ld n ), and the available mass (mass avail,n ).
z) Select the consequence area mitigation reduction factor (fact mit ) from Table 14. aa) b For each release hole size, calculate the energy efficiency correction factor, (eneff n ) using equation below.
-15 (20) bb) Determine the fluid type, either TYPE 0 or TYPE 1 from