<®

Avail able online at mwv.sciencedirect.com

Structural Integrity

* f^* ■

CrossMark

Procedia Structural Integrity 2 (2016) 3654-3659

JLI U^LUI Ul <1 PLtyi !l.y

ScienceDirect P rOCed Í 0

www.elsevier.com/locate/procedia

21st European Conference on Fracture, ECF21, 20-24 June 2016, Catania, Italy

Ri sk Based Approach to Integrity Assessment of a Large Spherical

Pressure Vessel

Aleksandar Sedmaka, Snezana Kirinb, Tamara Golubovica, Slobodan Mitrovic1; Petar

S tanojevicb

"Faculty of Mechanical Engineering, University of Belgrade, Kraljice Marije 16, 11000 Belgrade, Serbia bInnovation Center ofthe Faculty of Mechanical Engineering, Kraljice Marije 16, 11000 Belgrade, Serbia cEleotropower Industry of Serbia, EPS, Serbia

Abstract

The risk based approach has been applied, in its simplest form, i.e. by using the risk matrix to illustrate how the water proof test can shift risk; from high to very high level in the case of large spherical pressure vessel (ammonla storage tank;). Having in mind the basic definition ofrisk, being the product of the probability and consequence, and fixing the consequence at the highest level, only pronefiity ofunfavourable event (leakage and/or failure) has been evqluated. Toward this end, the failute assessment diagram (FAD) has been used here aa enothet simple eneinerring tool to estimate probability oa the failure, as the !unction om tine; position of the operating po!nt, i.ee defining prolmbility as the rotio between the deistimi; e ofthe operating point from thc zero pfmt, and the oppropriate distance between flue point on the limiting curve and zero point. This simple engineering tool to assess ^^c^riel mtegrity showed cleariy that water proof test is nmt alwayu recommended, because it disregareg possible stoble growth of urauks, which might rej^<;h critical size for" unstable growth, i.e. it does not prove that eailure will nds happenin auture imdor thc aame conditieng.

Copyright © 2d6 The Aulhorr. Publiched ley Eiseviet B.V. Thi s is ao open access article undrr fee CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Peer-review under re spac(ibility of tin; Stientific ^omi^it^ee of ECF21 .

Keywords/Risk based approach; ftiactaiat ietafiity affaffmaet; large pressure vaffat

1. Introduction

Risk; based approach is usually explained by the risk matrix, Fig. 1, using the simple definition of risk (product of probability and consequence). In the case of large pressure vessels, containing anmeonia, this is even simpler tosk, since the coeesequence category is certairiy the highest, thus reducing riskassessment to the probability category. Anyhow, there is still n quee^on if one use the nimple option (probability, which can tie defined using previous experience, e.g. as the number of events in certain periodof time, divided by the total number of pressure f^ssees operating in the same period of time) or more complicated one (e.g. API procedure, (American Petroleum Institute,

Copyright © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Peer review under responsibility of the Scientific Committee of ECF21.

10.1016/j.prostr.2016.06.454

2014, API 581, 2010), or its European competitor, RIMAP, Jovanovic (2011), both based on empirical rules). The first one is definitely not an option here, since it is oversimplified and has no relevance to any concrete problem, whereas the later one is based on experience, and it complexity does not necessarily leads to the correct prediction of probability. Therefore, yet another way to estimate the probability shall be applied here, based on fracture mechanics principles and structural integrity assessment, used to modify risk matrix approach.

Table 1.

Consequence category

A BCD E

1 Medium risk Very high risk

Probability category 2 3 4 5 Very low risk High risk Medium risk Low risk High risk Medium risk

One should keep in mind that the most critical part of a pressure vessel is welded joint, Sedmak (1996). As the case study, leakage of large spherical tank, used for storage of ammonia, will be analysed here. It was caused by undetected micro-cracks in welded joint, which have grown through the thickness during proof testing (cold-water test with pressure up to 50% above the operating pressure), Sedmak (2011). The testing of storage tanks before and after inspection has clearly shown the adverse effect of proof test in service, since it has indicated large number of new cracks in the locations of "old" ones. The macroscopic view of a typical through crack causing leakage is given in

Fig. 1. Macroscopic view of a typical through crack

Nevertheless, the full scale tests of welded pressurized equipment are the most informative when safety is considered, Sedmak (2011). In some cases they are inevitable despite high cost because they can give realistic answers relating the service behavior of welded joints. Hydrostatic pressure proof test can be classified as the full scale test. Hydrostatic pressure for proof test is often calculated using the formula p, = 1.3p, where p, is proof test and pris the design pressure. The logic behind this approach is that once a pressure vessel has withstood proof test, it will be safe in the exploitation under design pressure. Anyhow, there is a controversy behind this logic, because the proof test has provoked cracking and leakage, in number of cases, Sedmak (2011). Therefore, one of the main aims here is to show, even graphically, a detrimental effect of proof test on pressure vessel safety.

2. Risk Based Approach

The Extensive European project RIMAP, from 2001 until 2004, was introduced to offer a European standard for risk based management, including inspection, maintenance and control, Jovanovic (2011). It has produced four industry specific workbooks (petrochemical, chemical, steel and power generation industries), aimed to provide more specific guidance on how to apply the RIMAP approach. However, this approach is too complex, and will not be considered here. Instead, we present here only the risk matrix approach, as illustrated in Tab. 1. This approach uses well-known definition of risk being the product of the probability and the consequence.

In the risk matrix shown in Tab. 1, consequences are categorized, based on several parameters (health, safety, environment, business, security) as A to E; A indicates low, almost negligible consequences, and E refers to fatal and serious consequences. Probability categories are graduated 1 to 5, starting with very unlikely event, let day once in over a 100 years (1x10-4), ending with highly probable event occurring at least once in a year (1x10-1), Table 1. This is obviously oversimplified and somewhat arbitrary approach, as opposed to the complex ones, as defined in API and RIMAP documents. Anyhow, the concept of using risk matrix can be useful in combination with fracture mechanics approach and structural integrity assessment, as will be shown in the following text using large spherical storage tank as the case study.

3. Structural integrity assessment

In-service behavior of many structural components revealed that cracks lead to the fatal failure. One possibility to prevent such a scenario is to use Failure Assessment Diagramme (FAD) which provides analysis for a cracked component, in the scope of its structural integrity assessment. The basic concept is to evaluate ratios between the stress intensity factor and fracture toughness (Y coordinate), which can be interpreted as the probability of brittle fracture, and between the local stress and its critical value (X coordinate), which can be interpreted as the probability of plastic collapse, Fig. 2. The point defined by these two coordinates is either in the safe or in the unsafe region, which are separated by the limit curve obtained by applying Dugdale's plastic zone concept. Probability of failure can be estimated in the same way, based on the distance from the point to the corresponding point at the limit curve.

Fig. 2. Failure Assessment diagram

4. Case study - Large Spherical storage tank for ammonia

The analysis was performed on the spherical storage tanks for ammonia storage (volume 1000 m3, diameter _D=12500 mm and wall thickness t=25 mm, Fig. 3, Sedmak (2011)). The operating pressure wasp=6 bar and proof test pressure p=10 bar was applied together with non-destructive testing (NDT). The tanks have been constructed using

the microalloyed steel St.E460, (yield strength Rpo.2=480 MPa, ultimate tensile strength Rm=680 MPa, elongation .4.5=28%). Welding of St.E460 turned out to be much more complicated than anticipated, causing a lot of problems regarding cracking and leakage. There have been many investigations of this problem, including testing of fracture toughness, focused on welded joints and their different regions, especially the heat-affected zone (HAZ). Based on results of such testing, we have adopted here K^=2750 MPaVmm, as the minimum value for fracture toughness in HAZ.

Fig. 3. The spherical tank for ammonia storage

Different NDT methods (ultrasonic, dye penetrant and magnetic particles) have been used to test welded joints. The longitudinal cracks were considered as more dangerous due to their size (length up to 100 mm, depth up to 5 mm) and position (HAZ), Sedmak (2011). Macroscopic view of the crack is shown in Fig. 4.

Fig. 4. Macroscopic view of the crack

In order to evaluate its significance, the crack is presented as an edge crack with length equal to its depth (5 mm), schematically shown in Fig. 5, as if it was along the whole circumference.

Fig. 5. Schematic view of the crack

Therefore, the conservative approach has been applied, with the following data:

• PV geometry (thickness t=25 mm, diameter D=12500 mm);

• St.E460 steel: Reh=480 MPa, Rm=680 MPa; K^=2750 MPaVmm;

• crack geometry (edge crack, length 5 mm, ratio length/thickness=0.2);

• loading (max. pressure p=0.6 MPa, stress a=p-R/2-t=75 MPa, residual stress ctr=196 MPa - max. value transverse to the weld, no measurements available, no post weld heat treatment, 40% of the Yield Stress, [5]);

• curvature effect is negligible (t/R=25/12500«0.002).

• The SIF is calculated from: Ki=1.12<pR/2t+ctr)vra=(75+196)v^5=1075 MPavmm, leading to the ratio Kr=Ki/KiC=1075/2750=0.39.

The net stress is ctn=1.25-pR/2t,coefficient 1.25=25/20 due to the reduced cross-section (crack length 5 mm vs. thickness 25 mm), ctf=(ReH+RM)/2=580 MPa; Sr=(1.25-75)/580=0.16, the coordinates (Kr, Sr)=(0.39, 0.16). If one takes the ratio of distance from zero point to this point to zero point and distance between the zero point and the cross-section point on the limit curve, the result is 0.395, which can be taken as the probability of failure.

Now, the same calculation for the proof testing (pressure p=1 MPa) leads to the following results:KR=Ki/Kic=1288/2750=0.47, SR=ctn/ctf=0.27; the coordinates (0.47, 0.27) and the ratio 0.4.

The FAD is shown in Fig. 6, indicating these two pressure levels, 6 bar (design) and 10 bar (proof test), indicating detrimental role of the proof pressure.

Fig. 6. The FAD for two pressure values

Finally, one should consider the option of such an analysis which does not take into residual stresses. in that case, these two points have the following coordinates: (0.12, 0.16) for pressure 6 bar, and (0.20, 0.27) for pressure 10 bar, leading to the following probabilities of failure: 18.2% for pressure 6 bar, and 30.3% for pressure 10 bar. in this case, the probability of failure is simply proportional to the level of pressure.

5. Conclusions

Based on the results shown her, one can state the following:

• Risk based approach can be useful tool for assessment of structural integrity, even if using simple graphical presentation, i.e. the risk matrix.

• Basic structural integrity tools, such as FAD, can be used in combination with the risk based approach to show detrimental effect of proof test in the case of large spherical storage tanks.

• Detrimental effect of proof test is even more pronounced if one does not take into accounts the effect of residual stresses.

References

Standard API 510 Pressure Vessel Inspection Code: In-Service Inspection, Rating, Repair, and Alteration, American Petroleum Institute, 2014. API 581, The standard for quantitative Risk Based Inspection, American Petroleum Institute, 2010.

Jovanovic, A., Kauer, R., Renner, M., 2011. The "story" of RIMAP: from a research project over CWA to EN and global market.. .and iNTeg-Risk

(FP7)!, Seminar Standardisation in Research and Innovation, Brussels. Sedmak, S. Petrovski, B. Sedmak, A., 1996. Crack significance evaluation with special reference to welded structures, Materials Science 32, 231243.

Sedmak, A., Sedmak, S., Milovic, Lj., 2011. Pressure Equipment Integrity Assessment by Elastic-Plastic Fracture Mechanics Methods, monograph published by DIVK, Belgrade, Serbia.