relief valve case study

17

Click here to load reader

Upload: vuongthuan

Post on 01-Jan-2017

213 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Relief Valve Case Study

Unsteady State Relief Valve Evaluation

External Pool Fire Scenario

ByRame Sulaiman

Process EngineerProcess Engineering Associates, LLC

Copyright © 2009 Process Engineering Associates, LLC. All rights reserved.

www.ProcessEngr.com

Page 2: Relief Valve Case Study

Description of ProblemThe following slides describe a special case encountered while

performing a refinery relief valve evaluation project. A pool fire scenario is near a large bank of stacked horizontal heat exchangers used in a chiller unit. The solvent-oil mixture has a large boiling point variation. Also, these heat exchangers have a small liquid volume relative to their overall physical cylindrical volume size.

In this case Process Engineering Associates recognized that the normal API 520 and 521 standards would not accurately represent reality for a fire case relief scenario.

www.ProcessEngr.com

Page 3: Relief Valve Case Study

API Recommended Practice 520 & 521(Design Pressure > 15 psig)

Total Heat Absorbed (Btu/hr) from the fire− Q = 21,000*F*A0.82

F: environmental factor (insulation, earth covered, below grade storage...)A: wetted surface area (ft2)

Fire Relief Requirement (lb/hr) = Q / H− H: latent heat of vaporization (Btu/lb)

ASME Section VIII, Division 1Accumulated pressure limited to 121% of MAWP

www.ProcessEngr.com

Page 4: Relief Valve Case Study

API RP 521: External Pool Fire

Pool fire area – 2500 ft2 (30 ft radius)Wetted surface area to a height of 25ft above the source of flame

25 ft

R = 30ft

H

Layout Top View Side View of Fire Source

Chiller Unit

www.ProcessEngr.com

Page 5: Relief Valve Case Study

Description of Chiller to be Evaluated

Chiller – 12 double pipe exchangers− 10” outer pipe, 8” inner pipe, 42ft long− Arrangement: 2 columns x 6 rows

20” spacing between rows− No steel enclosure or insulation jacketing− Shell side fluid: Wax Free Oil and Solvent

80 vol% Solvent (MEK & Toluene)− Relief valve: 1D2 (0.11 in2 orifice area) @ 400 psig

www.ProcessEngr.com

Page 6: Relief Valve Case Study

Chiller Layout

Hliquid

Cross Section of Individual Exchanger (liquid in annulus)

10” Outer Shell

8” Inner Shell

20ft

Chiller HX Stack End View

www.ProcessEngr.com

Page 7: Relief Valve Case Study

API CalculationRelieving pressure (121% MAWP) = 484 psigTotal Surface area (12 exchangers) = 1418.4 ft2

Environmental credits – none, F =1 Q = 21000* 1 * 1418.40.82

= 8.07 MMBtu/hrLatent Heat of Vaporization, H = 67.5 Btu/lbFire Relief Requirement = 119,526 lb/hr (33.2 lb/s)Current relief valve (1D2) capacity is 6207 lb/hrRecommended valve size – 4L6 (2.853 in2)

www.ProcessEngr.com

Page 8: Relief Valve Case Study

Alternative Relief Path

An alternative relief path was evaluated through the 6”process piping.The process line ties into the inert gas header via the Flash Tower, the Solvent Cooler, and the Solvent Recovery Drum. The MEK Unit has a dedicated cooling tower. The above equipment is located outside the boundary for an external pool fire around the chillers.

www.ProcessEngr.com

Page 9: Relief Valve Case Study

MEK Unit Layout

Chiller 2

Chiller 1

Cooling Tower

Flash TowerMAWP: 18 psig

Solvent Cooler

Solvent RecoveryMAWP: 16 psig

> 100ft

Inert Gas HolderMAWP 12” WC

www.ProcessEngr.com

Page 10: Relief Valve Case Study

Dynamic Simulation Needed

Large variation in boiling point between solvent and wax free oil. The annulus volume is relatively small compared to shell surface area, therefore the wetted shell height will decrease rapidly as the solvent boils off.Using dynamic simulation the change in the rate of vaporization can be calculated as the liquid composition changes and the heat input decreases due to reduce liquid height.

www.ProcessEngr.com

Page 11: Relief Valve Case Study

Unsteady State SimulationChemCAD Simulation

Sequence for Varying Heat Load (Q)1) Run Simulation2) Export Vapor and Remaining Liquid streams to Excel3) Calculate new liquid height

Obtain wetted surface areaObtain new heat absorbed, Q

4) Import Q to exchanger5) Repeat steps 1 - 4

www.ProcessEngr.com

Page 12: Relief Valve Case Study

Unsteady State: Constant Heat Load

Graph of Vapor Rate and Remaining Liquid For E-3 Fire Scenari(Values per Exchanger)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

0 100 200 300 400 500

Time (s)

Mas

s (lb

) & T

emp

(F)

0

5

10

15

20

25

30

35

40

Vapo

r Rat

e (lb

/s)

Temp (F) Liquid Mass Vapor Rate

Peak solvent vapor rate = 12.8 lb/s (46,080 lb/hr) ‏

www.ProcessEngr.com

Page 13: Relief Valve Case Study

Unsteady State: Varying Heat Load

Graph of Vapor Rate and Remaining Liquid for E-3 Fire Scenario with

Variable Wetted Surface Area. (Values per Exchanger)

0

100

200

300

400

500

600

700

800

900

0 20 40 60 80 100 120

Time (s)

Mas

s (lb

) & T

emp

(F)

0

2

4

6

8

10

12

Vap

or R

ate

(lb/s

)

TempLiquid MassHeat InputVapor Rate

Peak WFO vapor rate: 9.97 lb/s (35,892 lb/hr) ‏

www.ProcessEngr.com

Page 14: Relief Valve Case Study

Vapor Recovery SimulationTo verify that the conservation valve on the Inert Gas Holder can handle the fire relief requirement, the vapor stream from the Unsteady State simulation is connected to the Solvent Cooler in the Vapor Recovery Simulation. Any uncondensed vapor then flows through the 3” pipe section that connects the Solvent Recovery Drum to the Inert Gas Holder, and the corresponding back pressure is calculated.

www.ProcessEngr.com

Page 15: Relief Valve Case Study

Vapor Recovery Simulation

www.ProcessEngr.com

Page 16: Relief Valve Case Study

Results Comparison API

Recommended valve size: 4L6 (2.853 in2)‏Maximum rate: 119,256 lb/hr Excessive back pressure process line, above MAWP of down stream equipment.

Unsteady StateRecommended valve size: 2J3 (1.287 in2)‏Maximum rate: 46,080 lb/hrAcceptable back pressure.Condenser can handle peak loads.

www.ProcessEngr.com

Page 17: Relief Valve Case Study

Shortfall of API 521 Method

In SummaryIt does not account for large variation in boiling points.It provides an unrealistic vapor relief rate when the exchanger volume is much smaller than the surface area.

www.ProcessEngr.com