Inherently Safe Heat Exchanger Network Design by Consequence Based Analysis

(1)

Procedia Engineering 148 ( 2016 ) 908 – 915

Peer-review under responsibility of the organizing committee of ICPEAM 2016 doi: 10.1016/j.proeng.2016.06.500

ScienceDirect

Available online at www.sciencedirect.com

4th International Conference on Process Engineering and Advanced Materials

Inherently Safe Heat Exchanger Network Design by Consequence

Based Analysis

Dzulkarnain Zaini

a

_{,Mohsin Pasha}

a,

_{*,Sunitaraj Kaura}

a

a_{Centre of Advance Process Safety (CAPS), Univerisity Technology Petronas Bandar Sri Iskandar, 31750 Perak, Malaysia}

Abstract

The failure of heat exchangers is an acute problem being experienced by chemical process industries. The loss of containment from this failure can cause enormous destruction and monetary loss in industries. The failure frequency of the heat exchanger network can be substantially improved through the inherently safer design. Moreover, integration between process design stages with risk and consequence estimation is extremely important in order to design inherently safe heat exchanger network. However, the lack of formal integration between process design stages with risk and consequence estimation results in the unproductive estimation of risk levels and consequences. Few studies on the integration of risk estimation with process design are available, but a viable framework is required to lay out the inherently safer heat exchanger network. Hence, based on the highlighted issue, application of integrated risk estimation tool (iRET) is proposed for designing the inherently safe heat exchanger network in the preliminary design stage.

Peer-review under responsibility of the organizing committee of ICPEAM 2016.

Keywords: Heat exchanger network; Inherently safer design; Process design silmulator; Preliminary design stage

1.Introduction

Inherently safer design (ISD) provides a coherent platform for removing the root cause of potential hazards. ISD strategies are different from the conventional safety techniques. In conventional methods, additional layers are added between the hazardous agent and people, environment and property to reduce the potential risk [1]. These

* Corresponding author. Tel.:+6-016-506-3706; fax:. +6-005-365-6176.

E-mail address:[email protected]

(2)

layers could potentially reduce the magnitude of hazards, but unable to eliminate the root cause of hazard. ISD focuses on the elimination of immediate cause of hazards rather than providing these complicated layers.

Nomenclature

CD Discharge coefficient E Explosion Energy

Ah Area of hole Hv Heating value

P Source Pressure Z Scaled distance ρ Density Ps Blast overpressure γ Ratio of heat capacities R Distance from source m Released rate Po Ambient Pressure

mf flammable released rate 'Ps Scaled Overpressure

Co Initial concentration V Causative Variable CLFL Concentration at lower flammable limit k1 & k2 Probit parameters

CUFL Concentration at upper flammable limit Y Probit variable

Process safety strategies are segregated in inherent, passive, active and procedural categories. Hazards can be significantly reduced by reducing the inventories of hazardous material and number of hazardous operations through inherent safety. In passive category, hazards is reduced by modifying the design features. Addition of protective layers such as safety interlocks, emergency shutdown system and control system are used for reducing hazards through active safety. Passive and active strategies are the sub classification of add-on or engineering safeguard [1].

Inherent and passive methodologies are considered more reliable than the rest of techniques. They depend on the physical and chemical properties of system rather than the proper and timely operation of safety devices [1]. In inherent safety, initiation cause of potential hazard is eliminated by altering the fundamental technology of process. Add-on protective layers cannot be able to eliminate the root of hazards. Moreover, these additional layers could potentially create complexities in process design and require repetitive maintenance and calibration.

ISD delivers an eminent platform to reduce the potential hazards. Possibilities of hazards could be significantly reduced by implementing ISD. Hence this safety design is deliberated as a primary prevention tool to avoid chemical accident [2]. ISD could be implemented throughout the lifecycle of the process system. However, opportunities to implement ISD strategies are become limited as the process design proceeds from initial to final stage as presented in Figure 1. Process design can be changed easily and economically at initial design stage. So, the best time to implement inherent safety is at early design stage [3, 4]. ISD also provides a gateway for developing an environmental benign process system. It could potentially develop a sustainable environment by reducing the fugitive emission rate [5].

Shell and tube heat exchangers (STHE) are being extensively deployed in the process industries. These heat exchangers are the component of integrated design process and serve as heater, cooler, condenser, evaporator, decomposer and boiler. The chemical process industries are being experienced an enormous failure of STHE. Loss of containment scenarios of these heat exchangers were also observed just after a few months of service [6, 7]. Various loss of containment scenarios such as small leak, continuous release and catastrophic rupture frequency of STHE is significantly higher as compared to other process equipments [8]. On 2nd April, 2010 an enormous catastrophic rupture of STHE was observed in Tesoro Anacortes Refinery Washington, United States [9]. Seven fatalities were reported in this accident. This tragedy was proclaimed as the largest fatal incident in US petroleum refinery since after the BP Texas city accident. The investigation report of this accident was issued by the US chemical safety board and it recommended that the failure of STHE could be significantly avoided by inherently safer design.

(3)

Integrated risk estimation tool (iRET) was designed to analyze the damage potential resulted from the explosion [10]. It is integrated with process design simulator for easily transferring of process information from simulator to MS Excel interface. MS Excel is used as a development platform in this framework. In this work, the same framework is implemented for the inherently safer design development of the heat exchanger network.

Fig. 1 Opportunities to implement ISD [3]

2.Inherently safer design framework for the heat exchanger network

Integrated risk estimation tool is modified for the inherently safer design of the heat exchanger network in this work. The conceptual framework is presented in Figure 2. The stepwise explanation of each is given below.

Fig. 2 Conceptual framework of designing the inherently safe heat exchanger network

(4)

The process flow diagram (PFD), material balance and process conditions would be required for the development of the process simulation diagram of the heat exchanger network. Aspen HYSYS V (8.0) is served as a simulation platform in this work, but user can integrate other process design simulators such as Aspen Plus and iCon. Process informations can easily be transferred from the process design simulator to MS excel interface via object linking and embedded (OLE) automation.

2.2 Flammability estimation

Flammability can be estimated by the difference of upper flammability limit (UFL) and lower flammability limit (LFL) respectively. UFL and LFL of the mixture can be estimated by the following equations.

¦

n 1 i i i mix UFL y 1 UFL (1)

¦

n 1 i i i mix LFL y 1 LFL (2)

2.3 Flammable mass estimation

Loss of containment scenario is considered under chock flow condition from each heat exchanger. The discharge rate under chock flow condition and flammable mass can be estimated by the following equations.

1 γ 1 γ h D 1 γ 2 γPρ A C m ¸¸ ¹ · ¨¨ © § (3) » » ¼ º « « ¬ ª ¸¸ ¹ · ¨¨ © § » » ¼ º « « ¬ ª ¸¸ ¹ · ¨¨ © § » » ¼ º « « ¬ ª ¸¸ ¹ · ¨¨ © § » » ¼ º « « ¬ ª ¸¸ ¹ · ¨¨ © § UFL 0 o UFL LFL 0 o LFL UFL 0 LFL 0 f C C ln π C 2C C C ln π C 2C C C ln erf C C ln erf m m (4) » » ¼ º « « ¬ ª ¸¸ ¹ · ¨¨ © § » » ¼ º « « ¬ ª ¸¸ ¹ · ¨¨ © § LFL 0 π 0 LFL LFL 0 f C C ln erf C 2C C C ln erf m m (5)

2.4 Estimation of explosion parameters

The TNO multi energy correlation are used to estimate the explosion parameters. The explosion energy can be estimated from the following equations.

f

v m

H

(5)

3 1 0 P E R Z ¸¸ ¹ · ¨¨ © § (7)

2

s Z 2.9456 Z 1.7095 1 Z 0.4462 0.6354 P Δ (8) 0 s s ΔP P P u (9)

2.5 Estimation of the potential damage

The fatalities, injuries and structure damage are considered as the possible consequent impact from the loss of containment of heat exchangers. The probit method can be implemented for analyzing these impacts [11]. The probit variable (Y) and probability (P) of these impacts can be computed from the following equations.

V ln k k Y ₁ ₂ (10) » » ¼ º « « ¬ ª ¸ ¸ ¹ · ¨ ¨ © § 2 5 Y erf 5 Y 5 Y 1 50 P (11)

Fatalities, injuries and structural damages were deliberated as the possible consequences of vapor cloud explosions. Probit parameters values for these impacts are listed in the Table 1. The probit variable value can be converted to probability by using Eq. (11).

Table1 Probit correlation values of various impacts

.

3.Results and discussions

The process flow diagram (PFD) of a typical ammonia synthesis loop heat exchanger network is illustrated as a case study. The original PFD of this heat exchanger network is presented in Figure 3. The Synthesis gas composed of hydrogen (H2), nitrogen (N2), ammonia (NH3), methane (CH4) and argon (Ar) is injected through the bottom of

ammonia convertor (4501) and yields about 15.0% ammonia conversion per pass. The product gas leaves from R-4501 at temperature 474.5 o_{C. A network of 08 heat exchangers from E-4501 to E-4508 is deployed for recovering}

the heat of reactor outlet gas. Eventually, this gas is cooled down to -4.0 o_{C and the liquid ammonia is separated}

from gaseous phase via two phase flash vessel (V-4501). The unconverted gas with makeup gas is recycled to the reactor. The makeup gas enriched of hydrogen gas is used to accomplish the deficit amount of reactant (H2 and N2)

consume in the reactor. The makeup gas is cooled in E-4509 before merging into the recycle stream.

The loss of containment scenario is considered from 25mm hole of each heat exchanger. The over pressure and consequent impacts resulted from the loss of containment of each heat exchanger are presented in Table 2 and 3 respectively. The distance from the source is kept constant during the estimation of explosion parameters and the potential damage.

Possible impacts Causative variable k1 k2

Death from lungs hemorrhage Ps -77.1 6.91

Ear drum rupture Ps -15.6 1.93

(6)

E-4509 is the hazardous heat exchanger in this network due to higher heating value of process gas. The higher release rate from this heat exchanger can cause serious injuries and enormous structure damage. Therefore, there is need to reduce the heating value of this gas by applying ISD strategies. Attenuation method is selected to reduce the heating value of a process gas in the modified heat exchanger network design. The modified design is presented in Figure 4. In this design, highly explosive process gas is mixed with low explosive gas before injecting into the heat exchanger (E-4509).

Fig. 3 PFD of original heat exchanger network.

Table 2 Estimation of explosion parameters in original heat exchanger network

Table 3 Estimation of consequent impacts in original heat exchanger network

T he expl osio n para met er Exchanger

Hole size Pressure Release Rate Release

duration Flammable Fraction

Flammable mass

Heating

value Over pressure

(mm) (Pa) (kg/sec) (sec) (kg) (kj/kg) (Pa)

E-4501 25 1.42E+07 8.64 300 0.41 1069.807 10992.32 15909.18 E-4502 25 1.42E+07 9.32 300 0.41 1153.830 10992.32 16549.19 E-4503 25 1.42E+07 11.16 300 0.62 2091.342 11359.88 23054.53 E-4504 25 1.42E+07 14.35 300 0.41 1776.958 10992.32 20776.32 E-4505 25 1.42E+07 13.25 300 0.62 2481.808 11359.88 25255.58 E-4506 25 1.42E+07 15.15 300 0.41 1876.033 10992.32 21383.23 E-4507 25 1.42E+07 13.81 300 0.62 2586.318 11359.88 25816.85 E-4508 25 1.42E+07 14.01 300 0.53 2231.192 12271.09 24864.41 E-4509 25 1.42E+07 8.21 300 0.86 2115.676 20262.21 31558.21 Exchanger

Probit value Probability

Fatalities Injuries Structure Damage Fatalities Injuries Structure Damage

E-4501 -13.15 3.07 4.45 0 2.69 29.12 E-4502 -12.89 3.15 4.57 0 3.20 33.18 E-4503 -10.70 3.79 5.53 0 11.28 70.31 E-4504 -11.39 3.59 5.23 0 7.89 59.07 E-4505 -10.10 3.96 5.80 0 15.01 78.80 E-4506 -11.20 3.64 5.31 0 8.74 62.30 E-4507 -9.95 4.01 5.86 0 16.02 80.61 E-4508 -10.20 3.93 5.75 0 14.32 77.45 E-4509 -8.62 4.39 6.45 0 27.23 92.65

(7)

and the potential damage estimation in the modified heat exchanger network design are presented in Table 4 and 5 respectively. In this design, the heating value of E-4509 process gas is reduced 40%, as a results consequents impact of injuries and structural damage has also been significantly reduced. Now this heat exchanger can be considered for design. However, critical safety review of E-4509 would also be required during the later design stages.

Fig. 4 PFD of modified heat exchanger network

Table 4 Estimation of explosion parameters in modified heat exchanger network

Table 5 Estimation of consequent impacts in modified heat exchanger network Exchanger

Hole size Pressure Release Rate Release

duration Flammable _Fraction

Flammable mass

Heating

value Over pressure

(mm) (Pa) (kg/sec) (sec) (kg) (kj/kg) (Pa)

E-4501 25 1.42E+07 8.64 300 0.41 1069.807 10992.32 15909.18 E-4502 25 1.42E+07 9.32 300 0.41 1153.830 10992.32 16549.19 E-4503 25 1.42E+07 11.16 300 0.62 2091.342 11359.88 23054.53 E-4504 25 1.42E+07 14.35 300 0.41 1776.958 10992.32 20776.32 E-4505 25 1.42E+07 13.25 300 0.62 2481.808 11359.88 25255.58 E-4506 25 1.42E+07 15.15 300 0.41 1876.033 10992.32 21383.23 E-4507 25 1.42E+07 13.81 300 0.62 2586.318 11359.88 25816.85 E-4508 25 1.42E+07 14.01 300 0.53 2231.192 12271.09 24864.41 E-4509 25 1.42E+07 13.71 300 0.53 2180.04 12271.09 24559.06 Exchanger

Probit value Probability

Fatalities Injuries Structure Damage Fatalities Injuries Structure Damage

E-4501 -13.15 3.07 4.45 0 2.69 29.12 E-4502 -12.89 3.15 4.57 0 3.20 33.18 E-4503 -10.70 3.79 5.53 0 11.28 70.31 E-4504 -11.39 3.59 5.23 0 7.89 59.07 E-4505 -10.10 3.96 5.80 0 15.01 78.80 E-4506 -11.20 3.64 5.31 0 8.74 62.30 E-4507 -9.95 4.01 5.86 0 16.02 80.61 E-4508 -10.20 3.93 5.75 0 14.32 77.46 E-4509 -10.28 3.91 5.72 0 13.79 76.36

(8)

4.Conclusion

A prototype iRET tool was used to study the failure potential assessment of heat exchanger network in the preliminary design stage. Initially, iRET was limited for piping systems only. Through this study, the scope of iRET has been extended for the heat exchanger network. The hazardous heat exchanger can easily be determined through the estimation of explosion parameters and the consequent impact analysis. Moreover, ISD strategies can easily be implemented to reduce the failure potential of the hazardous heat exchanger. Finally, this prototype has a potential to commercialize due to its easy understandable approach and widespread application in process industry.

Acknowledgements

The authors would like to thank Universiti Teknologi PETRONAS, Malaysia for providing short term internal research fund (STIRF) 0153AA-D13 that makes this project feasible.

References

[1] Faisal I. Khan and P.R. Amyotte, How to Make Inherent Safety Practice a Reality. The Canadian Journal of Chemical Engineering, 2003. 81(February ).

[2] Nicholas Askounes Ashford and G. Zwetsloot, Encouraging inherently safer production in European firms: a report from the field. Journal of Hazardous Materials, 2000. 78: p. 123–144.

[3] Hurme, M. and M. Rahman, Implementing inherent safety throughout process lifecycle. Journal of Loss Prevention in the Process Industries, 2005. 18(4): p. 238-244.

[4] Hendershot, D.C., An overview of inherently safer design. Process safety progress, 2006. 25(2): p. 98-107. [5] Hassim, M.H., et al., Fugitive emissions in chemical processes: The assessment and prevention based on

inherent and add-on approaches. Journal of Loss Prevention in the Process Industries, 2012. 25(5): p. 820-829. [6] Allahkaram, S., P. Zakersafaee, and S. Haghgoo, Failure analysis of heat exchanger tubes of four gas coolers.

Engineering Failure Analysis, 2011. 18(3): p. 1108-1114.

[7] Mousavian, R.T., et al., Failure analysis of a shell and tube oil cooler. Engineering Failure Analysis, 2011.

18(1): p. 202-211.

[8] Tugnoli, A., V. Cozzani, and G. Landucci, A consequence based approach to the quantitative assessment of inherent safety. AIChE journal, 2007. 53(12): p. 3171-3182.

[9] CSB. Catastrophic rupture of heat exchanger (Seven Fatalities) Tesoro Anacortes Refinery Anacortes,Washington. 2014; Available from: http://www.csb.gov/assets/1/19/tesoro_anacortes_2014-jan-29_draft_for_public_comment.pdf.

[10] Mohd Shariff, A., et al., Inherent safety tool for explosion consequences study. Journal of Loss Prevention in the Process Industries, 2006. 19(5): p. 409-418.

[11] Crowl, D.A. and J.F. Louvar, Chemical process safety: fundamentals with applications. 2001: Pearson Education.