A

p
p
lic
a
tio
n
S
o
lu
tio
n
Summary
Dynamic simulation can be
used to simulate a typical
distillation column or
fractionator relief scenario
to calculate the column
relief load more accurately
than using conventional
methods, while
maintaining an acceptable
margin of safety.
Business Value
Eliminate the need for
additional relief valves
on unit revamp or
debottlenecking projects
Eliminate or reduce the
cost for additional flare
system piping for refinery
expansion projects
Increase throughput
through units that are
constrained by relief
capacity
INTRODUCTION
Dynamic simulation can simulate a typical distillation column or fractionator relief
scenario (such as cooling water failure, partial power failure, and total power failure).
The objective of the analysis is to calculate the column relief load more accurately
than using conventional methods while maintaining an acceptable margin of
safety. The peak relief fow may then be used to size relief valves or used within an
overall fare analysis to determine relief valve back pressures. In most cases, the
use of dynamic simulation can reduce the need for additional relief valves or fare
system expansion. This example discusses how dynamic simulation can reduce the
calculated relief loads by 25% or more as compared with conventional calculations.
DYNAMIC SIMULATION BENEFITS SUMMARY
Eliminate the need for additional relief valves on unit revamp or debottlenecking
projects
Eliminate or reduce the cost for additional fare system piping for refnery
expansion projects
Increase throughput on units that are constrained by relief capacity
DYNAMIC SIMULATION APPROACH
Conventional methods for calculating column relief loads are very conservative
and can lead to over-designed relief systems, unnecessary fare, or fare header
replacements during unit revamp and debottlenecking projects.
Dynamic simulation provides an alternative to conventional calculation methods and
provides substantially lower relief loads. For fractionation towers with wide boiling
ranges, signifcant fare load reductions are possible. Dynamic simulation accounts
for the limited inventory of light components and the sensible heat required to boil
off the heavier components.
Dynamic simulation is consistent with the approach of API-521. API RP 521 Section
3.3 (1995) has two components:
1. For mixtures, light components are relieved before heavier components.
2. Change in vapor rate and molecular weights at various time intervals should be
investigated.
Dynamic simulation provides a safe and documented alternative to conventional relief
load calculations methods that result in substantially lower calculated relief loads.
Dynamic Simulation Example:
Column Relief Load Analysis
CONVENTIONAL APPROACH
The conventional approach used by many process designers is to use a heat balance method around
the entire column. If condenser cooling is lost, the heat balance is recalculated accordingly and the
net heat input is applied to a tray at or near the top of the column. The heat of vaporization is used to
calculate relief fow based on the following equation:
Where:
W - Relief fow
f - Feed or product fow
h - Feed or product enthalpy
Q - Reboiler or condenser duty
- Heat of vaporization
DYNAMIC SIMULATION MODEL
Diagram 1 depicts a distillation column, which illustrates some of the principles involved in using
dynamic simulation for column relief load analysis.
Diagram 1: Column Relief Load
The following assumptions were made in the development of this simulation model;
Tower
The DYNSIM

Tower Model includes both vapor and liquid holdup on each tray.
Tray liquid holdup assumes clear liquid height. This is a conservative assumption as it increases the
amount of volatile low boiling point components.
The Tower has the same number of theoretical stages as the steady state simulation. The inventory of
each theoretical stage is adjusted by the ratio of actual trays to theoretical stages.
A DYNSIM Separator simulates the column sump. The Separator includes an internal weir to simulate
the thermosyphon reboiler baffe. Accurately modeling the sump inventory is important as reboiler
heat transfer rate reduces as the inventory is depleted.
Pumps and Valves
DYNSIM rigorously models pumps, valves, and controllers. However, there are simplifed models
available for situations where not all the equipment data is available. For example, the DYNSIM
StreamSet model simulates the overhead liquid product level control valve and refux fow control valve.
A DYNSIM Valve and PID simulate the feed fow control valve and fow controller. The rigorous valve
model calculates the reduced feed fow to the column that may occur as the column pressure rises.
Condenser
The overhead condenser is modeled with an equation to reduce surface area when the overhead
accumulator foods.
Relief Valve
The DYNSIM Relief Valves are set to modulate rather than pop open. A modulating relief valve provides
the peak relief rate while holding the column pressure at a nearly constant rate as required to determine
the relief load. This is the preferred approach when determining a peak relief fow to use in an overall
fare system back pressure analysis or to size relief valves for a new column. Alternatively, the relief valve
can be pop acting based on the available relief valve orifce area to determine if pressure rises above
the allowable accumulation (10% for single valves and 16% for multiple valves).
Controls
Per API Recommended Practice 521, controllers fail in their last positions unless their automatic
action increases the relief load. Many controllers are not required to be modeled and the simulation
does not need to match the control confguration used in the actual plant.
Refux cascade fow controller operates to keep a constant mass fow rate. The master controllers are
assumed to not change the set point the slave controllers.
The overhead pressure controller locks in position since its action to open on high pressure would
reduce the relief valve load.
Tray temperature controllers on reboiler steam are not required to be simulated since a temperature
increase during relief would reduce reboiler heat transfer.
SCENARIO CASES AND RESULTS
Table 1 summarizes the three relief scenarios and provides the results for the conventional calculation
and peak DYNSIM relief fow. Table 2 shows the actual scenarios as implemented in the DYNSIM
Scenario utility.
Table 1: Scenario Cases and Results
Table 2: DYNSIM Scenarios
DISCUSSION OF RESULTS
For all cases, the equipment failure event occurs at the fve-minute timeframe.
Case 1 Loss of Cooling Water
Figure 1 shows the relief rate for the loss of cooling water case. The peak relief rate is 42,000 kg/hr,
24% lower than the conventional calculation. Refux continues to empty the contents of the overhead
receiver into the column and cascades down through the trays to the reboiler where it is vaporized to
increase pressure and create relief.
Figure 1: Case 1 - Loss of Cooling Water
Case 2 Loss of Reux/Overhead Pump
Figure 2 shows the relief rate for loss of the Refux Overhead Product Pump. The peak relief rate is
120,000 kg/hr, 40% lower than the conventional calculation. The pressure initially decays as the refux
stops the supply of volatile material entering the column. However, the overhead receiver foods in an
approximate four minutes, which eliminates condenser cooling, and the pressure begins to increase to
create relief. There is a large spike at 23 minutes as the low boiling material reaches the reboiler.
Figure 2: Case 2 - Refux Failure
Invensys, the Invensys logo, ArchestrA, Avantis, Eurotherm, Foxboro, IMServ, InFusion, SimSci-Esscor, Triconex, and Wonderware are trademarks of Invensys plc, its subsidiaries or affliates. All other
brands and product names may be the trademarks or service marks of their representative owners.
2010 Invensys Systems, Inc. All rights reserved. No part of the material protected by this copyright may be reproduced or utilized in any form or by any means, electronic or mechanical, including
photocopying, recording, broadcasting, or by any information storage and retrieval system, without permission in writing from Invensys Systems, Inc.
Invensys Operations Management 5601 Granite Parkway III, #1000, Plano, TX 75024 Tel: (469) 365-6400 Fax: (469) 365-6401 iom.invensys.com
Case 3 Loss of Reux / Overhead Pump and Feed
Figure 3 shows the relief rate for loss of the overhead pump. The peak relief rate is 15,000 kg/hr, 78%
lower than the conventional calculation. Pressure frst falls due to loss of refux. However, since there is
a loss of feed the reboiler is not supplied with low boiling material and the relief quickly subsides.
Figure 3: Case 2 - Refux and Feed Failure
Additional Work
It is the process engineers obligation to ensure that the results of a simulation model are appropriate
for the design. The following is a partial checklist to help validate a column relief study using dynamic
simulation.
Review electrical bus arrangements to see which electrical devices can trip at the same time. Loss of
multiple devices due to a common electrical supply is a typical relief scenario.
Check initial column tray holdup calculations to be sure that the inventory estimate is adequate. It is
allowable if you overestimate the actual inventory using a clear liquid approach.
Investigate pop-acting relief valve to determine maximum pressure using the actual relief system
confguration.
Be sure that the liquid levels are not low such that vapor is blowing through liquid control valves.
In many cases, conservative limiting case assumptions can be made to simplify simulation
development. Perform sensitivity runs to determine the impact of these assumptions on the results to
see if further simulation refnement is warranted. This approach is often far less time consuming than
attempting to create a more rigorous simulation.
Rel. 01/10