Brunet(Dynamic Simulation and Process Control)

DYNAMIC SIMULATION AND PROCESSCONTROL WITH ASPEN HYSYS

Robert BrunetPolitechnika Warszawska

1

Contents

1 INTRODUCTION TO DYNAMIC SIMULATION 31.1 Mathematical model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61.2 Holdup Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81.3 Nozzles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91.4 Distributed and Lumped models . . . . . . . . . . . . . . . . . . . . . . . . . 101.5 Static Head Contribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111.6 Heat Loss Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121.7 Integration Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131.8 Unit Operation Guidelines for Dynamics . . . . . . . . . . . . . . . . . . . . 161.9 Rating the equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171.10 Trouble Shooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

2 INTRODUCTION TO DYNAMIC SIMULATION 202.1 Resistance Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.2 Volume Balance Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3 FUNDAMENTALS OF PROCESS CONTROL 303.1 3.1. Level Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323.2 3.2. Choosing the Correct Control . . . . . . . . . . . . . . . . . . . . . . . . 353.3 Temperature Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403.4 The Process Reaction Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

4 NGL EXTRACTION PLANT 444.1 Steady-state NGL extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . 464.2 Moving from Steady-State to Dynamics . . . . . . . . . . . . . . . . . . . . . 494.3 Control Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

5 INLET SEPARATION PLANT 595.1 Steady State model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615.2 Preparing for Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645.3 Installing control strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 695.4 Completing the model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

2

1 INTRODUCTION TO DYNAMIC SIMULATION

The design and optimization of a chemical process involves the study of both steadystate and dynamic behaviour. Steady state models can perform steady state energy andmaterial balances and evaluate different plant scenarios. The design engineer can use steadystate simulation to optimize the process by reducing capital and equipment costs while max-imizing production.

With dynamic simulation, you can confirm that the plant can produce the desired productin a manner that is safe and easy to operate. By defining detailed equipment specificationsin the dynamic simulation, you can verify that the equipment functions as expected in anactual plant situation.

With dynamic simulation, you can investigate:

• Process optimization.

• Controller optimization.

• Safety evaluation.

• Transitions between operating conditions.

• Start-up/Shutdown conditions.

The dynamic mode shares the same physical property packages as the steady state model.However, the dynamic mode needs a different solver with a different set of conservation equa-tions to be solved.

The Steady State mode uses modular operations which are combined with a non-sequentialalgorithm. Information is processed as soon as it is supplied. The results of any calculationare automatically propagated throughout the flowsheet, both forwards and backwards.

In steady state simulation material, energy, and composition balances are consideredconstant in time. In addition all the specifications are considered equally. For example, atemperature specification can be replaced by a vapour fraction specification or a column’sproduct flow rate specification is replaced by a composition specification in the reboiler. Thesimulator can solve with either specification.

In Dynamic mode material, energy and composition balances are not considered con-stant, over time. The equations for material, energy, and composition balances include anadditional ”accumulation” term, which is differentiated with respect to time. Non-lineardifferential equations can be formulated to approximate the conservation principles; but ananalytical solution method does not exist. Therefore, numerical integration is used to de-termine the process behaviour at distinct time steps and the solver should be run after theaddition of any unit operation to the flowsheet.

3

Dynamic process simulation offers a high level of realism with regards to material flowthrough the simulation thanks to the use of dedicated solver to calculate the pressure-flowtype of equations. Pressure and flow equations are solved simultaneously in a pressure-flowsystem of equations at every integration step. Temperature and composition specificationsneed to be defined at every boundary feed stream entering the flowsheet. Temperatureand composition through the flowsheet are then calculated by means of the energy andcomposition balances in a modular sequential way for every existing unit operation.

4

Learning Objectives

The contents of the module will help users to understand all the mathematical and engineer-ing assumptions behind the dynamic engine of the process simulator. The following aspectsof the dynamic model are presented:

• Mathematical model: Material and Energy Balance and Implicit Euler algorithm.

• Holdup Model and Non equilibrium Flash.

• Nozzles location.

• Distributed and lumped Models.

• Static head contribution.

• Heat loss Model.

• Integration Strategy.

5

1.1 Mathematical model

The dynamic mass, component and energy balances are similar to the steady state balances,with the exception of the accumulation term. It is the accumulation term which allows theoutput variables from the system to vary with time.

Material and Energy Balance

For the simple case of perfectly mixed tank, the material balance is as follows:

d(ρVdt

) = Fin · ρin − Fout · ρoutWhereF = flowrateρ = densityV = volume

(1)

For a multi-component feed, the balance for component, would be as follows:

d(CiVdt

) = Fin · Ci,in − Fout · Ci,out

WhereC = concentration

(2)

Equations (1) and (2) are a simplification of the more rigorous equations used insidethe simulator which also considers other phenomena such as vaporization, reactions, densitychanges, etc... And the energy balance is as follows:

d(u+k+gz)·ρVdt

) = Qin −Q(out) +Win −W(out) + (h · F · ρ)in − (h · F · ρ)outWhereu = internalenergyperunitmassh = enthalpyperunitmassk = kineticenergyperunitmassgz = potentialenergyperunitmassQ = headaddedorlostW = shaftwork

(3)

The ODE Solver

The ordinary differential equations are solved by the simulator by using the Implicit Eulermethod which solves by an approximation rectangular integration.

dydt

= f(y) = y(t)−y(t−∆t)∆t

y(t) = y(t−∆t) + ∆t · f(y, t) (4)

The advantage of implicit methods, such as the one from equation (4), is that they areusually more stable for solving a system of stiff equation, meaning that a larger step size ∆t

6

can be used.

The integrator property view allows adjusting the time step value to increase the speedor stability of the model. The integrator property view is available by pressing CTRL+I orselecting it from the Simulation menu.

The smaller the time step, the more closely the calculated solution matches the analyticsolution. However, this gain in rigour is being paid by the additional calculation timerequired to simulate the same amount of elapsed real time. A reasonable compromise isachieved by using the largest possible step size, while maintaining an acceptable degree ofaccuracy without becoming unstable.

7

1.2 Holdup Model

Dynamic behaviour arises from the fact that many pieces of plant equipment have somesort of material inventory or holdup. Therefore, the impact of changes in composition, tem-perature, pressure or flow from an inlet stream to a vessel with volume (holdup) are notimmediately seen in the outlet stream. The hold-up model predicts how the holdup andoutlet streams of a piece of equipment, respond over time to input changes to the system.

Calculations included in the holdup model are:

• Material and energy accumulation.

• Adiabatic PH flash calculation for vapour composition and pressure effects in thevapour holdup.

• Heat transfer.

• Chemical reaction.

• Flash efficiencies for the modelling of non-equilibrium behaviour between the feedphases of the holdup.

• The placement of feed and product nozzles on the equipment has physical meaning inrelation to the holdup. For example, if the vapour product nozzle is placed below theliquid level in a separator, only liquid exits from the nozzle.

Calculation assumptions for the holdup model are:

• Each phase is assumed to be well mixed.

• Mass and heat transfer occur between feeds to the holdup and material already in theholdup.

• Mass and heat transfer occur between phases in the holdup.

In the real world, the extent of mixing the feeds with a holdup depends on the placementof the feed nozzles, the amount of holdup, and the geometry of the piece of equipment.In the simulator, you can indirectly specify the amount of mixing that occurs between thefeed phases and the existing holdup using feed, recycle, and product efficiencies. These feedefficiency parameters can be specified on the Efficiencies tab of the unit operation’s Advanceview.

8

1.3 Nozzles

In Dynamics mode, you can specify the feed and product nozzle locations and diameter forany piece of equipment. These nozzle placement parameters can be specified in the Holduppage under the Dynamics tab of the unit operation property view by pressing the Advancedbutton.

If you go to the Nozzles tab you can enter the nozzle’s elevation and diameter.

In Steady State mode, the bottom product stream of a vessel is considered to be at itsbubble point and the vapour stream at its dew point, unless user had specified some vapouroutlet pressure drop. In Dynamics mode, the vapour fraction of a product stream dependson the placement of feed and product nozzles on the equipment and the current liquid level:

For practical purposes, the simulator moves nozzles located at the extreme bottom ortop of the vessel very slightly. This minor adjustment is not displayed and does not impactthe static head contributions. The adjustment is mostly done so that users do not have toconsider nozzle diameters etc. carefully when setting elevations. The adjustment makes thecalculated values more realistic.

9

1.4 Distributed and Lumped models

Most chemical engineering systems have thermal or component concentration gradients inthree dimensions (x,y,z) as well as in time. This is known as a distributed system. If you wereto characterize such a system mathematically, you would obtain a set of partial differentialequations (PDEs).

If the x, y, and z gradients are ignored, the system is ”lumped”, and all physical prop-erties are considered to be equal in space. Only the time gradients are considered in suchan analysis. This consideration allows for the process to be described using ordinary dif-ferential equations (ODEs) which are much less rigorous than PDEs, thereby significantlysaving calculation time. For most instances, the lumped method gives a solution which is areasonable approximation of the distributed model solution.

The process simulator used in this course uses lumped models for all unit operations.Therefore, there are no thermal or concentration gradients present in a single phase and thetemperature and composition of each phase are the same throughout the entire hold up ofthe unit operation.

10

1.5 Static Head Contribution

If the simulator is using a Lumped Model approach, does this mean that there is no pressuregradient between the top and the bottoms of one equipment filled with liquid? The answeris Yes by default and No if you go to the Options Tab of the integrator and activate theEnable implicit static head contribution check box.

By selecting it, vessels can, optionally, be solved using implicit static head calculationsfor the pressure contributions associated with the liquid level inside the vessel, rather thanusing explicit static head calculations.

This option provides increased stability in applications where these static head contribu-tions play a crucial role. Static head is important in vessels with a certain level of liquid,because the liquid’s column exerts some pressure on the exit. For example, consider a verticalseparator unit operation that has a current liquid level of 50%. The static head contributionof the liquid holdup makes the pressure at the liquid outlet nozzle higher than that at thevapour outlet nozzle.

Nozzle location also becomes significant with this respect. The pressure flow relationshipfor the separator is different for a feed nozzle which is below the current liquid holdup levelas compared to a feed which is entering in the vapour region of the unit.

11

1.6 Heat Loss Model

The heat loss parameters can be specified for most unit operations in the Heat Loss pageunder the Rating tab. You can choose to neglect the heat loss calculation in the energy bal-ance by selecting the None radio button or select one of the two heat loss models available.The Simple model, allows you to either specify the heat loss directly or have the heat losscalculated from specified values.

The heat loss is calculated using the following equation:

Qout = U · A(Tf − Tamb) (5)

Where, UA is the overall heat transfer coefficient, Tf is the fluid temperature and Tamb

is the environmental temperature.The Detailed model allows you to specify more detailedheat transfer parameters.

The model assumes heat is lost or gained from the holdup fluid through the wall andinsulation by conductivity and convection. There are three radio buttons in the Heat LossParameters: the temperature profile, conduction and convention.

The temperature across the wall and insulation is assumed to be constant, and themodel considers a temperature profile composed by the fluid, wall, insulation, surroundingstemperature and a temperature gradient through the thickness of the wall and insulation,but temperature is constant along the vessel.

12

1.7 Integration Strategy

Because the pressure flow solver exclusively considers pressure flow balances in the network,pressure flow specifications are separated from temperature and composition specifications.Pressure flow specifications are input using the one P-F specification per flowsheet boundarystream rule.

While pressure and flow are calculated simultaneously in a pressure flow system of equa-tions, energy and composition balances are solved in a modular sequential fashion.

Furthermore, material, energy and composition balances in dynamic mode are not con-sidered at the same time. Material and pressure flow balances are solved for every timestep. But, energy and composition balances are defaulted to solve less frequently. Solvingmaterial and energy balance at every time step would be too computational expensive andis not required in most of the gas processing simulations.

The Steady State mode uses modular operations which are combined with a non-sequentialalgorithm. Information is processed as soon as it is supplied. The results of any calculationare automatically propagated throughout the flowsheet, both forwards and backwards. Indynamic mode, information is not processed immediately after being input. The integratorshould be run after the addition of any unit operation, or any other new piece of informationto the flowsheet.

Temperature and composition specifications, should be entered for every boundary feedstream entering the flowsheet. Temperature and composition are then calculated sequen-tially for each downstream unit operation and material stream using the holdup model.

You can access the Integrator view from the Simulation menu or by using the CTRL +I hot key:

13

The Integration Time group contains the following parameters:

• Units Units for the Current Time, End Time and Display Interval fields.

• Current Time Displays the time that the Integrator is running.

• Acceleration If running in Real Time, changing this field can speed up or slow downthe model by taking larger or smaller steps.

• End Time Allows you to specify the time at which the Integrator stops.

• Real Time Activates the Desired Real Time Factor field.

• Display Interval Visible only in Automatic Integration Control, this field containsthe time interval at which the simulator updates the views.

• Real time factorVisible only in Automatic Integration Control, this field is calculatedby dividing a time interval for a case by the actual time required to simulate that timeinterval. The Real time factor depends on the computer’s processing speed and thecomplexity of the simulation case.

14

The Integration Step Size group contains the unit for the integrator step size and the cellfor the step size value. By default is 0.5 seconds and while the integrator is running, thisvalue cannot be changed.

The Execution tab contains the parameters that indicate the frequency at which the dif-ferent balance equations are solved. The default values for Pressure flow equations, Controland Logic Ops, Energy Calculations, and Composition and Flash are 1, 2, 2, and 10 respec-tively. A value of 2 for the Energy Calculations means that an energy balance is performedevery 2 time steps.

• Pressure Flow Solver Since pressure and flow can change rapidly, their calculationsare solved at the highest frequency and should be left at its default, 1.

• Control and Logical Ops The default number should always be sufficient, but youcan reduce this number for special cases. (E.g., if you need rapid control responses orto mimic equipment where sample data can only be obtained at a low frequency.)

• Energy Calculations The energy calculation, interpolates between the flash calcula-tions. The value should be lower than that of the composition and flash calculations.

• Composition and Flash Calculations If you reduce this number, the flashes willbe performed more frequently. This can slow down the calculation speed, but it mayresult in more accurate results in some cases. This number can be reduced in caseswhere the phase change in an individual vessel is being studied and a high degree ofaccuracy is required with regard to the phase composition.

15

1.8 Unit Operation Guidelines for Dynamics

Before a transition from steady state to dynamic mode can be made, the simulation caseshould be set up so that a realistic pressure difference is accounted for across the plant. Somebasic steps you can take to set up a case in Steady State mode and then switch to dynamicmode are:

1. Identify material streams which are connected to two unit operations with no pressureflow relation and whose flow must be specified in Dynamic mode. These unit operations in-clude the separator operation and tray sections in a column operation. Add unit operations,such as valves, heat exchangers, and pumps, which define a pressure flow relation to thesestreams.

2. Size all the unit operations in the simulation using actual plant equipment or prede-fined sizing techniques.

3. Specify one pressure flow specification for each flowsheet boundary stream.

4. Identify key control loops that exist within the plant. Implementing control schemesincreases the realism and stability of the model. Disturbances in the plant can be modeledusing the Transfer Function operation.

16

1.9 Rating the equipment

Valves Rating information for the valve operation, including the valve type and Cv values,can be input on the Sizing page in the Rating tab. Valves should be sized using typical flowrates. The valve should be sized with a 50% valve opening and a pressure drop between 15and 100 kPa.

Mixer and Tee It is recommended to specify the mixer with the Equalize All optionin dynamic mode. With this, the pressure of the surrounding streams of the unit operationis equal if static head contributions are not considered. This is a realistic situation sincethe pressures of the streams entering and exiting a mixer or tee must be the same. It isalso recommended that, the dynamic tee model should not use the dynamic splits as spec-ifications, so that the flow to and from the tee is determined by pressures and resistancethrough the flowsheet. This is more realistic than using the split fractions which can alsocause complications with regards to flow reversal. These options are set on the Specs pageof the Dynamics tab in their respective operation views.

Pump and Compressors Rating information for the dynamic compressor, expander,and pump operations, can be input on the Curves and Inertia pages in the Rating tab. Ingeneral, two specifications should be selected in the Dynamics Specifications group, in theSpecs page of the Dynamics tab, in order for these unit operations to fully solve.

Heat Exchanger The dynamic heat exchanger can be specified as having a set pressuredrop or an Overall K-Value (pressure flow) relation. This option is set on the Specs page ofthe Dynamics tab in the heat exchanger property view: K-values can be calculated using theCalculate k button on the Specs page of the Dynamics tab in the operation’s property view.Heater and cooler operations are much like heat exchangers. However, they only have a sin-gle K-value on their process side. Be cautious of Heaters/Coolers with fixed duties. This cancause problems if the flow in the heater/cooler happens to fall to zero. It is recommendedto use a controller, or a Spreadsheet function, or a temperature specification to control thetemperature of a stream.

Separator Rating information including the volume of the vessel, boot capacity, andnozzle location can be entered on the Sizing and Nozzles pages in the Ratings tab. A sep-arator with no valves attached to the inlet and exit streams requires at most one pressurespecification. The other two streams are specified with flows. A more realistic way to runthe separator is to attach valves to the inlet and exit streams of the vessel. The boundarystreams of the separator with valves should be specified with pressure. Vessels (Separators,Condensers, Reboilers) should be sized for 5 - 15 minutes of liquid holdup time. Sizing andCosting calculations can also be performed using the Vessel Sizing Utility in the Sizing pageof the Rating tab.

Separation Columns For all separation columns, the tray section parameters includingthe tray diameter, weir length, weir height, and tray spacing, can be specified on the Sizingpage in the Rating tab of the Main TS property view. Tray Sizing can be accomplished for

17

separation columns using the Tray Sizing utility in the Utilities page. The trays are sizedaccording to the existing flow rates and the desired residence times in the tray. Any use ofTray Sizing should be restricted to Steady State mode Adjusting Column Pressure In steadystate, the pressure profile of the column is user specified. In dynamics, it is calculated usingdynamic hydraulic calculations. If the steady state pressure profile is very different from thecalculated pressure drop, there can be large upsets in flow in the column when the integratoris run. A reasonable estimate of the column’s pressure profile can be calculated using theTray Sizing utility. This utility provides a value in the Results tab. The column pressureprofile can be calculated using this value and a desired pressure specification anywhere onthe column. You can change the value to achieve a desired pressure profile across the column.This can easily be done by modifying the Weir height in the Rating tab in the Tray Sizingutility. Reducing the weir height lowers the static head contributions and lowers the value.In dynamic mode, the Nozzle Pressure Flow K-factors (found on the Dynamics tab of theMain TS property view) can also be adjusted to better model the pressure drop across thecolumn. Feed and product streams entering and exiting tray sections, should be at the samepressure as the tray section itself. Any large pressure differences between a feed or productstream and its corresponding tray section can result in large amounts of material movinginto or out of the column.

18

1.10 Trouble Shooting

Singular Problem This message indicates that not all of the equations in the pressure flowsolver matrix are independent of one another. This occurs when one or more equations areredundant.

For instance, if a valve operation is using a pressure drop specification, the inlet andexit streams cannot both be specified with pressure. The pressure drop equation becomesredundant. It is useful to over specify a singular problem. The simulator might be able toidentify the redundant pressure flow specification and allows the case to solve.

The PF Solver failed to converge This message indicates that one or more pressureflow specifications are unreasonable. This message can also appear if there are suddenlarge upsets to the simulation case. It is helpful to enter the Equation Summary Viewto identify problem areas in the flowsheet. Click the Full Analysis button (or PartitionedAnalysis button, if it is made available). By clicking the Update Sorted List button inthe Unconverged tab, the simulator shows the type of equation, location, and scaled errorassociated with the unconverged nodes in the flowsheet drop. Pay special attention to theunit operations with the largest errors in the Uncoverged tab. Check the vessel volumes ofthe uncoverged unit operations and ensure they are sized with reasonable residence times.Check the size of the valves attached to the unconverged unit operations.

19

2 INTRODUCTION TO DYNAMIC SIMULATION

Dynamic simulation adds more realism to the model thanks to the pressure flow solver(P-F Solver). This is a specific solving method to deal with pressure and flow related calcu-lations.

The P-F Solver considers the integration of pressure flow balances in the flowsheet. Thereare two basic equations, with pressure and flow as unique variables, which define most ofthese P-F variables:

• Resistance Equations - defining the flow between pressure holdups.

• Volume Balance Equations - defining the material balance at pressure holdups.

In general, the resistance equation calculates flow rates from the pressure differences ofthe surroundings piece of equipment. Resistance equations are derived from the Bernoulliequation and general friction factors for turbulent flow.

P1

g·ρ = P2

g·ρ + hf (6)

hf = fDLD

u2

2(7)

Substituting equation (2) in equation (1) and introducing the mass flow rate, results onequation (3), where all the variables that are not pressure or flow have been grouped in aflow conductivity constant which is the reciprocal of resistance to flow.

Flow = k√ρ ·∆P (8)

The pressure drop, ∆P, is the frictional pressure drop across the unit operation withoutstatic head contributions.

As shown, a resistance equation relates the pressure of two holdups and the flow thatexists between them. The following unit operations have a resistance equation associatedwith them: Heater, Cooler, Heat Exchanger, Air Cooler, LNG Heat Exchanger, Valves andColumn Trays. Following a similar approach, Pumps, Compressors and Expanders have aresistance equation where the heat flow and the pump or compressor work define the pressureflow relation of the unit operation.Valve flow CoefficientThe Cv is defined as the flow of water at 60Ao F in US Gallons per minute that passesthrough a control valve when the pressure drop is 1 psi and the valve is fully open.

Q(USGPM) = Cv · 1√ 1(psi)1(lb/ft3 (9)

20

Once the Cv of the valve is known, it is possible to estimate the flow when the valve isfully open for another pressure drop, by applying the equation:

Q = Cv√∆P

ρL(10)

By introducing the relative density of the liquid it is possible to calculate the flow atother conditions or for another fluid.

Q = Cv · f(x)√∆PρL

(11)

Finally, the inherent flow characteristic of a valve f(x) is introduced as the relation be-tween the tap position, x, and the product flow rate that flows through it as a fraction ofthe maximum flow rate. If the pressure drop, ∆P in the valve is constant, it is possible toestimate the resulting flow at other valve positions.

f(x) = FFmax

(12)

Volume Balance EquationIn dynamics mode, all unit operations with hold-up represent pressure nodes. There are unitoperations like the separator which contribute with only one pressure node and others, likethe column with multiple stages, which contribute with the same number of pressure nodesthan trays in the column.

During calculations in dynamics mode, the change in volume, V, of the total material(liquid and vapour phase) inside an equipment, is zero:

dVdt

= 0 (13)

For a fixed volume, the change in pressure node, P, is calculated as a function of thechange in temperature (enthalpy) and the change of accumulation within the equipment(hold up).

dPdt

= f(Temperature, flows, size) (14)

An increase in the feed flow rate, with a constant product flow rate, will result in theholdup increasing. The accumulation of vapour occupying a fixed volume will cause the nodepressure to rise. The liquid hold up causes an increase in liquid level which compresses thevapour holdup, causing the pressure to rise.

21

Learning Objectives

After the completion of this module users will be able to understand the foundations of thepressure flow solver:

• How the simulator calculates flow between nodes.

• How the simulator calculates the pressure at nodes.

• Sizing vessels and valves.

• The meaning of flowsheet boundary streams.

• How to define pressure flow specifications.

• How to define reverse flow condition.

22

2.1 Resistance Equations

In the process simulator, flows exiting from a material holdup are calculated from a volumebalance equation, specified by the user, or calculated from a resistance equation. In general,the resistance equation calculates flowrates from the pressure differences of the surroundingnodes. The simulator contains unit operations such as valves and heat exchangers whichcalculate flowrates using resistance equations. The resistance equations are modelled afterturbulent flow equations.

1.1.1. Create a new case. Use Water and Air as components. Select Peng Robinson (PR)as Property Package. Move to the Simulation Environment.

1.1.2. Add a new stream, Stream 1, at 25AoC and 4 bar with pure water as composition.Enter 1000 kg/h as mass flow rate.

1.1.3. Add a valve,VLV-100, with 3 bar as pressure drop and stream 2 as outlet stream.

1.1.4. What are mass flow and pressure for stream 2?If when building the simulation case we did input 1000 kg/h, this is the answer in Q1 for

the flow rate. It is then enough to calculate (P1 - deltaP) to know the pressure for stream2: 100 kPa. However, for answering the same question in dynamics mode it is necessary toapply the resistance equation for valves, which uses the valve flow coefficient.

The mass flow rate that passes through the valve is a function of the valve flow coefficient,Cv, and of the frictional pressure drop across the valve. Therefore, Q1 cannot be answeredin dynamics mode if we do not provide at least a value for the valve Cv if the size of thevalve is required.

1.1.5. Open VLV-100 property view and go to the Sizing page under the Rating tab.Enter 2 USGPM (US Gallon Per Minute) as Cv value.

1.1.6. Open Stream 1 property view; go to the Specs page of the Dynamics tab andensure that Pressure is the only active Dynamic Specification at 400 kPa.

1.1.7. The last step is the way to consider the pressure upstream of VLV-100 as a bound-ary condition of the integration algorithm.

23

1.1.8. Repeat step 1.1.6. for Stream 2.

1.1.9. Press the dynamic mode option and make the integrator active.

1.1.10. What is the mass flow and pressure for stream 2?

1.1.11. Open the VLV-100 property view and go to the Sizing page under the Ratingtab. On the Valve Operating Characteristics group you can see that a Linear Valve, 50%open, is selected by default. Change the Valve Opening from 50% to 100%.

1.1.12. What is now the mass flow rate?

The pressure flow matrix of the model has one equation and three variables: the twoboundary pressures and the mass flow rate. By defining two of them, the third one is calcu-lated by the equation.

In the example, we are specifying both boundary pressures and calculating the flow rate.But a valid alternative, could be to specify inlet pressure and flow, so we will have the outletpressure as calculated unknown variable.

1.1.13. Open Stream 1 property view and go to the Specs page under the Dynamicstab.

1.1.14. In addition to the already active Pressure specification, activate the Mass Flowone with the value you used to answer 1.1.10.

1.1.15. Then, for Stream 2, deactivate the Pressure specification.

1.1.16. What are the mass flow and pressure for stream 2?1.1.17. What are the mass flow and pressure for stream 2, if you close the

valve from 100% to 75%?

24

2.2 Volume Balance Equations

1.2.1. Add a separator to the flowsheet and connect stream 2 as inlet. Create a new stream3, for the Vapour Outlet, and stream 4 for the Liquid Outlet.

The vessel becomes red and a message appears in the property view Volume not specified.For simulating equipment with holdup in steady state mode, an underlying principle is thatthe physical volume of the equipment, and thus, the volume of material in the equipment atany time remain constant. This principle does not applies in dynamics mode and thereforeto run the model in dynamic mode we need to provide the volume of each equipment withholdup.

1.2.2. Go to Sizing page of the Rating Tab and enter a volume of 50 L.

In exercise 1, we specified pressures for stream 1, P1, and stream 2, P2, to calculate theflow, FV-100, across VLV-100:

F = Cv · f(50%)√P1 − P2 (15)

And for answering Q4 and Q5, we moved the stream 2 pressure specification, to thestream 1 flow specification. In this way, it was possible to calculate the pressure of stream2.

F = Cv · f(50%)√P1 − P2 (16)

Because V-100 is a perfectly mixed pressure node, streams 2, 3, 4 and the vessel itself, areall at the same pressure. However, to run the case in dynamics mode, we still need to addmore information. As mentioned above: under dynamics mode, the flow rate through anyunit operation depends on the pressure of the surrounding piece of equipment. For V-100,the surroundings and the equipment are at the same pressure. Therefore the simulator doesnot have enough information to calculate the flow for streams 4 and 5.

1.2.3. Add two valves, VLV-101 and VLV-102 and connect streams 3 and 4 as the inletsto each one of them.

1.2.4. Define streams 5 and 6 as the outlet streams.

1.2.5. Read the Cv of VLV-100 and enter the same value for VLV-101 and VLV-102.

25

All material streams within the simulator have to be solved for pressure and flow, andall holdup unit operations have to be solved for pressure (only). Therefore in the flowsheetthere are 13 variables to be solved, 6 streams, times 2 variables plus the vessel pressure. Ifwe consider that there is no accumulation in valves, we can assume the following relationswhich reduce the variables from 13 to 10:

FV LV−100 = F1 = F2

FV LV−101 = F3 = F4

FV LV−102 = F5 = F6

dPV−100/dt = f(T, FV LV−100, FV LV−101, FV LV L−102, size)

(17)

Therefore, to satisfy the degrees of freedom of this pressure-flow system of equations wemust input three additional equations or pressure flow specifications (7 - 4 = 3). With thepressure and flow specifications that we already have for stream 1, we are setting:

FV LV−100 = 1503kg/hP1 = 400KPa

(18)

Then, we still need to add one new pressure flow specification to solve the system.

1.2.6. Move the stream 1 flow specification to a pressure specification equal to 1 atm instream 5.

1.2.7. Add a pressure specification equal to 1 atm in Stream 6.

The pressure-flow system of this example will be solved according to the following equa-tions:

FV LV−100 = CvV LV−100f(50%)(P1 − PV−100)1/2

FV LV−101 = CvV LV−101f(50%)(PV−100 −P5)1/2

FV LV−102 = CvV LV−101f(50%)(PV−100 −P5)1/2

dPV−100/dt = f(T, FV LV−100, FV LV−101, FV LV L−102, size)

(19)

26

1.2.8. Run the simulator.

1.2.9. A new message appears: it indicates that, at process conditions, the fluid in theseparator is only liquid. This enters in conflict with the default specification of 50% liquidvolume level. Then the simulator asks the modeller what to do to initialize the liquid in thevessel:

1.2.10. Click on 100% Liquid option and allow the simulator to run for a pair of minutesof simulation time.

1.2.11. Press CTRL+D for accessing to the Databook and select the Variables tab.

1.2.12. Press the Insert... button and add the following variables.

1.2.13. Move to the Strip Charts tab.

27

1.2.14. Click the Add button to include a strip chart with default name DataLogger1in the list of available strip charts.

1.2.15. Click the Active checkbox for each variable that you want to display in this stripchart.

1.2.16. In the View group, click the Strip Chart... button to view the selected strip chartor just double-click the name of the strip chart you would like to view.

1.2.17. Run the integrator.

1.2.18. How much is the liquid percent level in V-100?1.2.19. What would you do to decrease it to 50%?1.2.20. If the feed valve, VLV-100, is fully closed and the bottoms valve, VLV-

102 is fully open, why the level is not decreasing?1.2.21. Go to the PFD and observe the pressures, (Shift+P). There is no pressure drop

between the vessel and the boundary streams.

1.2.22. Decrease stream 6 boundary spec pressure to 70 kPa.

1.2.23. Observe the mass flow rates, (Shift+M). Now, there is a positive flow rate at thevessel bottoms. However, you should observe as well that, the vessel pressure is lower thanthe boundary pressure specification for stream 5. This means that, reverse flow is, likely tohappen and that stream 5 is feeding the vessel through VLV-101.

1.2.24. What is the composition of this unexpected vessel feeding?

If it is wanted to drain the liquid out of the vessel, some other material has to substitutethe volume of the liquid being drained. In any real plant, there will be air from atmosphereor nitrogen from a blanket, feeding back the vessel or, if external feeding does not exist,

28

internal vacuum is then created.

1.2.25. Open stream 5 and go to the Dynamics Tab.

1.2.26. Press the Product Block... button.

1.2.27. In the Conditions tab, leave temperature as specified variable, and its defaultvalue, 25A◦C.

1.2.28. Go to the Composition tab and define reverse flow composition as pure air.

1.2.29. Run the integrator.

1.2.30. How much is the percent liquid level of V-100?

29

3 FUNDAMENTALS OF PROCESS CONTROL

In the second course module a vessel in dynamics mode was modeled but it was notpossible to control the liquid level. In this module the basics of Process Control will be cov-ered by the practical example of controlling the liquid level and the temperature of the vessel.

The PID Controller operation is the primary means of manipulating the model variablesin Dynamics mode. It adjusts a stream OP (output) to maintain a specific flowsheet variablePV (process variable) at a certain value SP (set point).

A variety of Feedback, Feedforward and other control schemes can be modeled by modi-fying the tuning parameters in the PID controller operation. Controller parameters can bemodified to incorporate proportional, integral and derivative action into the controller.

Besides the PID controller other five controller operations can be modeled:

• Split Range Controller Several manipulated variables are used to control a singleprocess variable. Here both manipulated variables are driven by the output of a sin-gle controller. However, the range of operation for the manipulated variables can beindependent of each other. Typical examples include the control of the pressure in achemical reactor by manipulating the inflow and outflow from the reactor.

• Ratio Controller In the Ratio Controller the objective is to keep the ratio of twovariables, the load and the manipulated, constant.

• MPC Controller The Model Predictive Control (MPC) controller addresses the prob-lem of controlling processes that are inherently multi-variable and interacting in nature,in other words, one or more inputs affects more than one output.

• DMC Plus Controller The DMCplus Controller engine runs in Aspen DMCplusOnline. You are required to have the DMCplus link, Aspen DMCplus Online, CIM-IO kernel and ACO Base licenses to run DMCplus in the Process simulator. WithDMCplus controller the Process Simulator works like a Real Plant.

• Profit Controller The Profit Controller (only available in UniSim Design) allows theuser to configure the models, create Profit Controller and run it in the simulation toevaluate the model and controller design...

• A Digital On/Off control operation is also available. The Process Variable (PV)that needs to be monitored and the output (OP) stream which is manipulated aredefined. When the PV reaches a specified Threshold value, the Digital Point eitherturns the OP On or Off, depending on how the Digital Point has been set up.

30

Learning Objectives

In this module we will cover an overview of the basics of Process Control Theory. Further-more, the user will learn how to implement a PID controller in the simulation model.

31

3.1 3.1. Level Control

3.1.1. Open the case you were working with in the first module, where it was not possibleto control the liquid level percentage of V-100 or just the starting case for this workshop.

3.1.2. Add a PID controller to the PFD and double click on its icon to get access to theConnections tab. The Connections tab allows you to select both, the Process Variable, PV,and the Manipulated Variable or Output Process for the controller, OP. It is comprised ofthe objects described in the table below:

3.1.3. Press the Select PV button and select Object V-100 and Variable Liquid PercentLevel. The Process Variable, or PV, is the variable that must be maintained, or controlledat a desired value.

3.1.4. Press the Select OP button and select Object VLV-102 and Variable ActuatorDesired Position (ADP).

3.1.5. Go to the Parameters tab. By default you get access to the Configuration page.The Configuration page allows you to set the process variable Range, controller Action, op-erating Mode, and depending on the mode, either the Set Point, SP, for the Process variableor the Operating Point, OP, for the Manipulated Variable, as well as the Tuning of thecontroller.

32

If we open VLV-102, positive ∆OP, the bottom output flow of the vessel increases andwe expect that the liquid percent level will decrease. Therefore, we have a negative steadystate gain and our action mode has to be Direct.

For the Controller to become operational, we must define the minimum and maximumvalues for the PV (the Controller cannot be switched from Off mode unless PVmin andPVmax are defined).

The process simulator converts the PV range into a 0-100% range, which is then usedin the solution algorithm. The following equation is used to translate a PV value into apercentage of the range:

PV (%) = PV−PVmin

PVmax−PVmin· 100 (20)

3.1.6. Select Direct Action.

3.1.7. Enter 0% and 100% as minimum and maximum PV values respectively.

3.1.8. Once you provide these values (as well as the Control Valve span), you may selectthe Automatic mode, and give a value for the Set point.

3.1.9. Switch the Mode to Auto.

33

3.1.10. Type 50% as Liquid Percent Level SP.

You can select where the signal from the controller is sent using the drop-down list inthe Execution field. If you select Internal, the controller confines signals generated to stayWithin the simulator. If you select External, the controller sends the signals to a DCS, if aDCS is connected to the process simulator.

3.1.11. Leave the Default Execution option, Internal.

34

3.2 3.2. Choosing the Correct Control

The Tuning group allows you to define the constants associated with the PID control equa-tion.

You should consider what type of performance criteria is required for the set point vari-ables, and what acceptable limits they must operate within. Generally, an effective closedloop system is expected to be stable and cause the process variable to ultimately attain avalue equal to the set point. The performance of the controller should be a reasonable com-promise between performance and robustness. A very tightly tuned or aggressive controller,gives good performance, but is not robust to process changes. It could go unstable if theprocess changes too much. A very sluggishly-tuned controller delivers poor performance, butis very robust. It is less likely to become unstable.

In the paragraphs below you will find some details of the different configurations thatcan be implemented for a feedback controller.

Digital On/Off

The most rudimentary form of regulatory control is on-off control. In the process simulator,it is implemented using the Digital Point operation. An excellent example of on-off controlis a home heating system. Whenever the temperature goes above the SP, the heating plantshuts off, OP=0%, and whenever the temperature drops below the SP, the heating plantturns on, OP=100%.

Since the controller cannot throttle the actuator, but only turn it on or off, the primarycharacteristic of on-off control is that the PV is always cycling about the SP.

The rate at which PV cycles and the deviation of PV from the set point are a functionof the system dynamics or dead time and capacity. The On/Off controller is an appropriatecontroller if the deviation from the set point is within an acceptable range and the cycling

35

does not destabilize the rest of the process.

Proportional Control

Proportional control is the simplest continuous control mode that can damp out oscillationsin the feedback control loop. This control mode normally stops the process variable, PV,from cycling but does not necessarily return it to the SP.

P-only control is implemented in the Process Simulator by setting the values of Td and Tito ¡empty¿ in the PID Controller operation. With P-only control, oscillations that occur inthe process variable due to disturbances or changes in the set point dampen out the quickest(have the smallest natural period) among all other simple feedback control schemes.

The output of the proportional control is defined as:

OP (t) = K · E(t) + b (21)

As it could be observed in the below figure, the larger is the controller gain, the loweris the error. However, increasing K makes the loop unstable. If K has a value such thatthe loop gain is equal to one, the loop will oscillate with a period that is a function of thenatural characteristics of the process and it is called the natural period.

In general Proportional control is suitable when a fast response to a disturbance is re-quired. However a sustained error occurs where the PV does not return to the set point evenwhen steady-state is reached. The sustained error is called offset and is undesirable in mostcases. Therefore it is necessary to eliminate offset by combining proportional control withthe integral control mode.

36

Proportional + Integral Control

The action of integral control is to remove any error that may exist. As long as there is anerror present, the output of this control mode continues to move the FCE.

OP (t) = OP (t0) +1TiE(t)dt (22)

When Proportional and Integral control are combined, oscillations can be dumped outand return the process variable to the set point. Under PI control, the gain has an effect notonly on the error, but also on the integral action.

OP (t) = K(OP (t0) +1TiE(t)dt) (23)

When we compare the equation for a PI controller to that for a P-only controller, we seethat the bias term in the P-only controller has been replaced by the integral term in the PIcontroller. Therefore, the integral action provides a bias that is automatically adjusted toeliminate any error.

Typically, the response period of PV under an I controller is much slower than for a Pcontroller, therefore because the addition of the P action, the response period under PI islonger than a P controller and shorter than an I controller. The integral time, Ti, is definedas the amount of time required for the controller output to move an amount equivalent to theerror. Because the relationship between Ti and the control action is reciprocal, increasing Tiresults in less integral action, while decreasing Ti results in greater integral action. Thereforeby increasing Ti we have less integral action and a shorter response and a behavior closerto only proportional controller. Thus, the integral time should be decreased just enough toreturn the process variable to the SP.

PI is the most common controller found in plants and it is suitable when offsets cannot betolerated. PI controller combines accuracy (no offset) with a relatively quick response time.

37

However, the added integral action acts as a destabilizing force which can cause oscillationsin the system and cause the control system to become unstable.

Proportional Integral Derivative Control

The purpose of derivative action is to provide lead to overcome lags in the loop. In otherwords, it anticipates where the process is going by looking at the rate of change of error.For D action, the output equals the derivative time, Td, multiplied by the derivative of theinput, which is the rate of change of the error. When PI is combined with the D action, thisone adds the additional response speed required to overcome the lag in the response fromthe integral action.

OP (t) = K(E(t) + 1TiE(t)dt+ TddE(t)

dt) (24)

The addition of the derivative mode in the PID controller provides a response similar tothat of a P-only controller but without the offset because of the integral action. Thereforea PID controller provides a tight dynamic response but, since it contains a derivative block,it cannot be used in any processes in which noise is anticipated.

The following is a list of typical controller tuning parameters appropriate for variousprocesses. There is no single correct way of tuning a controller. The objective of processcontrol is to provide a reasonable compromise between performance and robustness in theclosed loop response. These initial values are a kind of suggestion. They help to obtain tightcontrol. They can be later adjusted further if the closed loop response is not satisfactory.Tighter control and better performance can be achieved by increasing the proportional gain.Decreasing the controller gain would result in a slower, but more stable response.

38

Generally, proportional control can be considered the principal component of the con-troller equation. Integral and derivative action should be used to trim the proportionalresponse. Therefore, the controller gain should be tuned first with the integral and deriva-tive actions set to a minimum. If instability occurs, the controller gain should be adjustedfirst. Adjustments to the controller gain should be made gradually.

39

3.3 Temperature Control

3.3.1. Add an energy stream to your simulation, Q-100, and connect it to V-100.

3.3.2. Enter 0.0 kJ/h as heat flow value for this stream.

3.3.3. Add a PID controller.

3.3.4. Select Vessel Temperature of V-100 as Process Variable Source.

3.3.5. Select Q-100 as Output Target Object and Control Valve as its variable.

The Control Valve button appears if the Output Target Object is a material or energystream instead of a valve unit operation. By pressing it, we have access to the Flow ControlView, FCV.

The FCV that appears for an energy stream is dependent on the type of duty streamselected:

• Direct Q duty consists of a simple power value (in other words, BTU).

• Utility Fluid takes the duty from a utility fluid (in other words, steam) with knownproperties.

40

The type of Duty Source specified can be changed at any time by clicking the appropriateradio button in the Duty Source group.

The default option is Direct Q. In this view option, the SP appears, and you may specifythe minimum (Min. Available) and maximum (Max. Available) cooling or heating available.You need to define these values in a way that your OP works at around 50% under yoursteady state target conditions.

3.2.6. Enter 0.0 kJ/h as the minimum available duty and 2.0e+05 kJ/h as the maximumone.

3.2.7. Go to the Parameters option in the Configuration page and enter 0AoC and 100AoCas controlled variable range PV values.

3.2.8. Which has to be the Control Action for this controller?

3.2.9. Switch the controller to Manual Mode.

3.2.10. Create a New chart. Go to Tools/Databook and select Strip Charts tab. Pressthe Add button.

3.2.11. Open the property view of the Temperature Controller and go to the Configura-tion page. Drag and drop the SP, PV and OP values to the new strip Chart.

3.2.12. Run the integrator until you get the steady state behaviour.

41

42

3.4 The Process Reaction Curve

The process reaction curve is probably the most widely used method for identifying dynamicmodels. It’s simple to perform and although it is the least general method, it providesadequate models for many applications and it is very helpful to tune PID controllers fromgeneral guidelines.

The process reaction curve method involves the following four actions:

• Allow the process to reach steady state.

• Introduce a single step change in the input variable.

• Collect Input and output response data until the process reaches again steady state.

• Perform the graphical process reaction curve calculations.

Change the OP of TIC-100 from 50% to 75% and run the integrator to the new steadystate. You should get a response like the one in the figure.

The graphical calculations determine the parameters for a first order with dead timemodel represented by the following equation:

Y (s)X(s)

= Kp · e−θsτ · s+ 1

WhereY (s)istheoutputofthesystemX(s)istheinputofthesystemKpisthesteady − stategainτistheT imeconstantofthesystemθisthedeadtomeoftheprocess

(25)

43

4 NGL EXTRACTION PLANT

Natural Gas Liquids (NGL) consist of hydrocarbon components in a produced gasstream which can be extracted and sold in their respective market. NGL extraction istypically justified:

• To meet a gas sales specification requirement such as a hydrocarbon dewpoint or

• to upgrade the market value of the produced gas and liquid streams.

Condensation processes are the most widely used processes for the extraction of NGLfrom natural gas. In this workshop we will build a NGL extraction process by Mechanical re-frigeration. Mechanical refrigeration plants utilize a commercial refrigerant such as propaneor R-22 to chill the gas. Process temperatures are seldom less than about -40A◦C. Thisprocess is used both for hydrocarbon dewpoint control and NGL sales.

The workshop is divided in three exercises. In the first one, the process model is built insteady state. In the second one, we cover all the aspects for the transition from steady stateto dynamics mode: 1) adding resistance unit operations, 2) rating all the equipment and 3)defining pressure boundary streams. Finally the third exercise is focused on the addition ofcontrollers and strip charts to follow the process dynamics.

44

Learning Objectives

After completion of the module, users would have had the opportunity to develop the dy-namic model of an mechanically refrigerated plant to extract the Natural Gas Liquids (NGL)of a gas mixture.

Users will be able to size equipment, define pressure flow specifications and add stripcharts and controllers.

45

4.1 Steady-state NGL extraction

The gas produced in the inlet separation section is dehydrated and sent to the NGL Sepa-ration section. In this case, an external refrigerated process will be modelled. The thermo-dynamic method and the first unit operations will be incorporated now in the simulationmodel:

The process and equipment data are given in the following tables:

46

47

Continue to build the chiller plant. Figure below shows the process schematic for thisadditional part:

The new equipment data and operating conditions are provided in the below table:

48

4.2 Moving from Steady-State to Dynamics

Adding unit operations

4.2.1. Identify material streams which connect two unit operations with no pressure flowrelation. Notice that boundary streams are coming from holdups or going to holdups. There-fore, if they are connected to a hold-up unit operation, a new unit operations which definea pressure flow relation, such as valves, heat exchangers and pumps, have to be added tothese streams. It is also possible to specify a flow specification on these streams instead ofusing an operation to define the flow rate.

4.2.2. Which streams in this flowsheet connect two unit operations with nopressure flow relation and will need an additional pressure flow relation or itsflow must be specified?

4.2.3. Add a new valve and name it VLV-103.

4.2.4. Connect Stream 11 as inlet of VLV-103. Create an outlet stream and name it 12.Enter 1800 kPa as outlet pressure.

4.2.5. Add a new valve and name it VLV-102.

4.2.6. Connect Stream 7 as inlet of VLV-102. Create an outlet stream and name it 17.Enter 3900 kPa as outlet pressure.

4.2.7. Add a new valve and name it VLV-101.

4.2.8. Connect Stream Inlet Gas as outlet of VLV-101. Create an inlet stream and nameit 1. Enter 1550 kPa as inlet pressure.

4.2.9. Equipment sizing. All unit operations in the simulation need to be sized using thedimension of the actual plant equipment or by using pre-defined sizing techniques.

49

Sizing the Valves

4.2.10. Open VLV-101 property view and go to the Specs page of the Dynamics tab.

4.2.11. Under the Dynamic Specifications group, there are two possible dynamic specifi-cations you can choose to characterize the Valve operation. If the Total Delta P checkboxis activated, a constant pressure drop is assumed across the Valve operation. With thisspecification, the flow and the pressure of either the inlet or exit stream must be specifiedor calculated from other operations in the flowsheet. The flow through the Valve is notdependent on the pressure drop across the Valve.

4.2.12. If the Pressure Flow Relation checkbox is activated, the flow rate through theValve is calculated from the valve equation, so will be the pressure of the streams enteringand exiting the Valve. If this option is selected the user must provide the valve size.

4.2.13. If the user does not have this information (valve’s Cv), it is possible to size thevalve on the Sizing page in the Ratings tab.

4.2.14. For sizing a valve the following information is required:

• Valve operating characteristics group (Linear, Quick Opening Equal Percentage ourUser Table).

• Stream Conditions (Normal Valve Opening, Pressure Drop, Mass Flow Stream Com-position, Inlet Pressure and Temperature).

• Sizing Method (Cv and Cg).

4.2.15. The sizing calculation method is the same for all valve manufacturers and types,with the exception of the Simple Resistance Equation. All valve manufacturers and types

50

have Cv and Cg methods to calculate the flow rate. The difference between the manufactur-ers and types is the equations and constants used to calculate the flow rate within the valve.

4.2.16. A valve should be sized using typical flow rates with a 50% valve opening and apressure drop between 15 and 100 kPa.

Sizing the Separators

4.2.17. In the Geometry group under the Rating Sizing page , you can specify the vesselorientation, shape, and volume. The geometry of the vessel is important in determining theliquid height in the vessel. There are four possible vessel shapes: flat cylinder, sphere, hori-zontal ellipsoidal cylinder, horizontal hemispherical cylinder. The liquid height in a verticalcylindrical vessel varies linearly with the liquid volume. There is a nonlinear relationshipbetween the liquid height, and the liquid volume in horizontal cylindrical and spherical ves-sels.

4.2.18. Open D-1 Scrubber property view and go to Sizing Rating page. Select Verticalorientation for a Flat Cylinder and press Quick Size button.

4.2.19. The Quick size option is sizing the vertical vessel for L/D ratio equal 3.0 and 5minutes of residence time. Vessels (Separators, Condensers, Reboilers) should be sized for 5- 15 minutes of liquid holdup time. Sizing could be also performed using the Vessel Sizingutility.

Sizing the Heat Exchangers

4.2.20. The dynamic heat exchanger can be specified as having a set pressure drop or aOverall k-Value (pressure-flow) relation. This option is set on the Specs page of the Dynam-ics tab in the heat exchanger property view: K-values can be calculated using the CalculateK button on the Specs page of the Dynamics tab in the operation’s property view.

4.2.21. Open E-100 property view and go to Dynamics Specs page. Press the Calculatek button and activate the Overall k dynamic Specification.

4.2.22. Enter 0.5 m3 as Volume.

4.2.23. Special caution has to be taken with Heaters/Coolers with fixed duties. This cancause problems if the flow in the heater/cooler happens to fall to zero. It is recommendedto use a controller, or a Spreadsheet function, or a temperature specification to control thetemperature of a stream.

4.2.24. Select Product Temp Spec within Model Details sub window and enter 45 AoCas Product Temp specification.

51

4.2.25. Repeat the last three steps for E-102 with a Product Temp Spec equal to -25AoC.

4.2.26. The heat exchanger model in Dynamics shares information with the DynamicRating mode. Therefore, it is good practice to converge the steady state flowsheet withDynamic Rating Mode before moving to Dynamic Simulation.

4.2.27. For solving E-101 with the dynamic rating mode you need a recycle operation forStream 10.

4.2.28. Add a recycle and a new stream as outlet, 10 bis. Connect 10 bis to the HXinstead of stream 10.

4.2.29. Open E-101 property view and go to the Model page of the Dynamics tab. En-ter 0.5 m3 and 3 m3 for the Tube and Shell volumes and 1.0e+5 kJ/C-h as Overall UA value.

52

4.2.30. Go to the Dynamics Specs page and press the Calculate K’s button.

4.2.31. Go to the Design Parameters page and switch from the Weighted model to theDynamic Rating model.

4.2.32. Go to the Dynamics Specs page and switch from Delta P Specification to k spec-ification in the shell and tube cells.

53

4.2.33. Because the Dynamic rating model calculate pressure drops from the Dynamic kSpecification, the Delta P values in the Parameters page of the Design tab must be deletedto avoid inconsistencies.

Sizing the Compressor and the Pump

4.2.34. In general, two specifications should be selected in the Dynamics Specifications groupin the Specs page of the Dynamics tab in order for Compressors, Expanders and Pumps tofully solve. Efficiency is one of the recommended specs and either Head or Pressure rise isthe second one. If available, compressor pump curves make excellent specifications.

4.2.35. You should be aware of specifications, which cause complications or singularityin the pressure flow matrix. Some examples of such cases are:

• The Pressure rise checkbox should not be checked if the inlet and exit stream pressuresare specified.

• The Speed checkbox should not be checked if the Use Characteristic Curves checkboxis unchecked.

• Duty specs on pumps can only be used if a recycle and pressure control to protect thepump are modelled. Otherwise use a pressure rise spec.

4.2.36. Go to the Specs page of the pump Dynamics tab and select 62% efficiency andpower as dynamic specifications.

4.2.37. Go to the Specs page of the compressor Dynamics tab and select 72% efficiencyand duty as dynamic specifications.

Boundary Stream Specification

4.2.38. Pay special attention to equipment with fixed pressure drops. Any fixed pressuredrop specifications in equipment can yield unrealistic results, such as flow occurring in thedirection of increasing pressure.

4.2.39. The last specification that needs to be included before the case can be moved tothe dynamic mode is to add Pressure specifications to all Boundary Streams.

4.2.40. Go to the top right corner palette and modify the code of colours from DefaultColour Scheme to Dynamic P/F Specs.

4.2.41. Add the following Boundary Specs.

54

55

4.3 Control Strategy

4.3.1. Press the Dynamics mode.

4.3.2. Run the Solver for a few minutes.

Feed Flow Control

4.3.3. Add flow controllers according to the following information.

4.3.4. Add a level controller according with the following information.

4.3.5. Add a temperature controller according with the following information.

56

4.3.6. Go to the Model Details in the Specs Dynamics page of E-102 and switch fromProduct Temp Spec to Supplied Duty.

4.3.7. Press [CTRL+F] to get access to the Face Plates window.

4.3.8. Select LIC-102 and TIC-102 and press the Open button.

4.3.9. Run the integrator and observe if the SP is achieved.

4.3.10. Add a Level controller according with the following information.

57

58

5 INLET SEPARATION PLANT

A lean gas condensate enters a separation plant where the different fluids will be sep-arated for further processing. The purpose of this separation in stages is to reduce thepressure on the produced fluids in steps to maximize each phase’s recovery. If the separatorpressure is too high, large amounts of light components will remain in the liquid phase atthe separator and will be lost along with other valuable components to the gas phase in thestock tank. Conversely, if the pressure is too low, large amounts of light components willbe separated from the liquid, and they will take with them substantial quantities of inter-mediates and heavier components. Consequently, it is necessary to optimize the separatorspressure in winter and summer seasons. Considerable gains can be realized by performingprocess simulations to optimize the separator pressure.

Knockout drums, 2 and 3-phase separators, valves, coolers, heaters, pumps and compres-sors of a typical Oil & Gas separation plant are added to the model in steady state.

Later, additional information is added to the model to adapt it for dynamic simulation:sizing information, pressure drop relationships, and boundary conditions will be included ina stepwise manner to systematically understand each step.

A control strategy is implemented to have the system working in closed loop. PID andonoff controllers are going to be used in this Module.

59

Learning Objectives

The content of the module will help users to understand how to install, connect andsolve unit operations first in steady state mode and later in how to move that model to adynamic operation. The guided exercises will show how to install unit operations of fourdifferent types:

• Simplified heat exchange devices: Cooler, Heater.

• Flash Separators: 2 & 3-phase separators.

• Piping equipment: Valves, mixers.

• Rotating equipment: Pump, compressors.

At the end of the module students will have learned how to use unit operations in dynamicsimulation models to calculate stream conditions and power requirements to accomplishcertain process working conditions. In later modules these unit operations will be definedwith a higher degree of rigorousness.

60

5.1 Steady State model

Through a guided exercise, the below flowsheet will be modelled in steady state mode toobtain initial results for moving it later to the dynamic environment.

5.1.1. Create a New Case and install a Fluid Package, with a Component List containingthe following Traditional components: N2,CO2, C1, C2, C3, iC4, nC4, iC5, nC5, nC6 andH2O. And the following Hypothetical components:

5.1.2. Use the Peng-Robinson equation of state as Property Package.

5.1.3. Create a Material Stream with the conditions in Table below.

61

5.1.4. Install a valve downstream of Produced Fluids (that will be later used for flowcontrol) and specify an outlet pressure of 64.8 bar. Name that valve Manifold.

5.1.5. Install a 2-phase separator to split the stream into one gas and one mixed-liquidstream. Name it HP Separator.

5.1.6. The HP Separator vapour stream enters an exchanger to decrease its temperaturedown to 27 oC.

5.1.7. Install a Cooler to perform such simulation. Consider that a 50 kPa pressure droptakes place in the cooler.

5.1.8. The 2-liquid phase stream leaving the HP Separator is flashed through a valvebefore entering the intermediate pressure separator.

5.1.9. Install a valve and fix its outlet pressure at 23.1 bar.

5.1.10. Install a 3-phase separator and name it IP Separator. Use the valve outletproduct as the feed for this separator.

62

5.1.11. The light liquid product stream is flashed again in a valve and warmed up tocompensate the Joule-Thompson effect.

5.1.12. Install a valve, then a heater with a 50 kPa pressure drop, and specify the outletheater pressure to be 5.5 bar and the temperature, 85 oC.

5.1.13. Install now the low pressure separator (LP Separator) to split the resulting threephases.

5.1.14. Install the pump by double-clicking its icon in the Object Palette and specify itsworking conditions to be 70% adiabatic efficiency to obtain an outlet pressure of 70 bar.

5.1.15. Place again a valve (50 kPa pressure drop) and a Cooler (50 kPa pressure drop,60 C outlet T) downstream of the light liquids stream.

5.1.16. Install valves downstream of the rejected water streams in IP and LP separators.Collect all water streams using a Mixer, which has to Equalize All pressures, to obtain afinal water stream at 4 bar pressure.

63

5.2 Preparing for Dynamics

The size and the physical characteristics of every piece of equipment will affect the way thevariables of the model respond to the time evolution of the dynamic model. Key aspects ofa transient response like residence time, capacitance, resistance, etc. are directly related tothe equipment size.

Before being able to start a run of a simulation in dynamic mode, all these values need tobe incorporated into the model. Using realistic values will additionally ensure the stabilityof the model and the accuracy of the results obtained.

There are two main aspects to focus on when adding the physical dimensions of the unitoperations: volume, that affects the pressure in the holdup and the residence time, and flowresistance that affects the transfer of material from one unit to another one. So, as seen inprevious modules, when entering this phase of dynamic model construction, the user needsto ensure that proper volumes are supplied for every unit operation and reasonable pressuredrop relationships are defined in every unit to account for the pressure gradient across thewhole simulation model.

In its current status the model contains 3 vessels, 3 heat transfer devices, 1 pump, 6 valvesand a mixer. Consequently, it is necessary to provide rating information for all these devices.

The software allows for a scalable approach to equipment sizing, from simply just enteringthe vessel volume (and the rest of dimensions are set to default values according to internalrules) to a complete access to all rating parameters. For the sake of simplicity and due tothe introductory level of this module, the simplest equipment sizing approach is being usedhere. Ask your instructor if you are curious about the additional capabilities of the software.

Most of the sizing information is going to be added in the Rating tab of every unit.Another tab that contains sizing information is the Dynamics tab. Using both tabs, the usershould be able to enter all the dimensions of the equipment.

5.2.1. Open the HP Separator property view and move to its Rating tab/Sizing page.

5.2.2. Enter a volume of 50 m3 and allow the simulator to auto-calculate the rest ofdimensions. Leave the Geometry at its defaults (vertical, flat cylinder).

64

5.2.3. The vessels do not offer any resistance to flow and in rating them, it is not neces-sary to worry about pressure drop correlations.

5.2.4. Size the IP and LP separators as horizontal, flat cylinders, of 30 m3 and 25 m3respectively.

5.2.5. For the IP Separator move to the Dynamics tab and see that in the Specs pageyou can find the same rating information. Just to start from a different initial value, fix theLiq Volume Percent at 40% instead of the default 50%.

5.2.6. For the three separators, once in the Rating tab, move to the Nozzles page. Checkthe elevations (both, in meters and in height %) at which the software has placed the differentnozzles by default. See that the boot is taken into account when measuring the elevationsof the nozzles.

5.2.7. The nozzle elevations are important values to remember because they will deter-mine what mixture is leaving through each nozzle.

5.2.8. The volumes of the three heat transfer devices are considered to be equal to 5 m3each. In these unit operations, the volume value is added on the Specs page of the Dynamicstab. In every heat transfer device, move to that page and enter the desired volume (See that

65

a default value of 0.1 m3 is already there, too small to be realistic).

5.2.9. Still in the Model Details group, move as well from a Supplied Duty model to aProduct Temp Spec model by switching the radio button. Ensure that the Product Tempshown is the one you specified when building the model.

5.2.10. The heat transfer devices do offer resistance to flow. The way the simulator willcalculate this resistance is by using a k constant in the turbulent flow equation. This k canbe calculated from the equivalent length of the installed equipment or can be back-calculatedby the software using the pressure drop and flow values used in the development of the steadystate model.

5.2.11. Still in the same Specs page of the Dynamics tab there is another group, DynamicSpecifications, where the user can use the Calculate k button to estimate the value of k anda check box to activate its use during dynamic simulations.

5.2.12. Repeat the procedure for every heat transfer device.

For the most realistic dynamic simulation of a centrifugal pump, the pump performancecurves are necessary. However, if process conditions remain stable, an acceptable simulationcan be achieved by using a fixed efficiency and a specified duty. As pump performance curvesare not available, the second approach will be followed in this exercise.

5.2.13. Open the Pump property view and go to Specs page of the Dynamics tab. Ensurethat the Efficiency and the Power are the two active (checked on) specifications in the list. Ifnecessary, this value of Power can be later modified by the user during dynamic simulationruns, if the flow is significantly different from the one in Steady State.

66

5.2.14. Open the Manifold valve and move to its Rating tab/Sizing page:

See that the Cv and the Cg values are missing but that a Size Valve button is available.When the button is used, the simulator uses the information that appears in the SizingConditions group plus the Valve Operating Characteristics to determine a current value ofthe valve’s Cv.

5.2.15. Press the Size Valve button for the Manifold valve.

5.2.16. Move to the Dynamics tab, Specs page and ensure that in the Dynamic Specifi-cations group the Pressure Flow Relation is selected.

5.2.17. Repeat these steps for the rest of valves in the model.

The Dynamic model should now be almost ready to run. It only misses the boundaryconditions between which to integrate the differential equations. The number of necessaryboundary conditions normally coincides with the degrees of freedom of the pressure-flowsystem of equations, which in turn usually coincides with the number of boundary streams(feeds and products) in the model. The most realistic setup of the model that can be pre-pared is to use pressure boundary conditions and allow the model to run due to the pressure

67

gradient that these conditions establish.

5.2.18. Switch the PFD Colour Scheme (you find it on the PFD top right corner) fromDefault to Dynamic P/F Specs.

5.2.19. Identify the boundary streams in the model: Produced Fluids, ProducedWater, Condensate Export, Gas Cooler Out, IP Vap and LP Vap.

5.2.20. Remembering the colour scheme mentioned in Module 1, check what boundaryspecifications are already active in the model.

5.2.21. Modify the existing specifications or create new ones in order to have only apressure specification in every boundary stream.

5.2.22. Check theDynamics Assistant to see if there is any potential trouble remaining.

5.2.23. Move to the Dynamics environment.

5.2.24. Run the integrator by pressing the green traffic light icon for a few simulationminutes to check if the integrator is capable of starting the model for dynamic simulation.

The simulation is now running in open-loop, without any control action to maintain thevariables at desired set points. Of course, a dynamic simulation needs as well as any realplant a control strategy to avoid undesired runaway situations due to upsets.

68

5.3 Installing control strategy

The simulation case is in Steady State environment and ready to have the control strategyadded. Let’s begin supplying the flow control of the system, by adding a control of the flowof produced fluids.

A flow controller

5.3.1. From the Object Palette, select a PID controller and place it on the PFD, close enoughto the Manifold valve.

5.3.2. The Connections page of a PID controller allows the user to select the ProcessVariable (PV), the variable that is necessary to maintain at a certain desired value (SetPoint, SP); and the Output Variable (OP), the variable that is going to be manipulated tohave the PV achieve the SP.

5.3.3. For this flow controller select the Molar Flow of Produced Fluids stream asPV and the Actuator Desired Position of Manifold valve as the OP:

As discussed in previous modules, a flow controller acting on a valve needs a Reverseaction. If the Steady State Environment results are going to be the normal operating pointfor Dynamic simulation, it is good practice to use a PV Range of twice the SS value (from 0flowrate to double the SS result). Good starting values for the Tuning parameters of a Flow

69

controller (in preparation for later fine tuning) can be Kc = 0.1 and Ti = 0.25 minutes.

5.3.4. Implement the above guidelines to define the flow controller:

A level controller

Now each vessel needs to have the level controlled to ensure that the mixtures leaving it areof the desired type. Let’s begin by defining a liquid level controller for the HP separator.

5.3.5. Install again a PID controller close enough to the HP Separator. Connect it byselecting as PV the Liquid Percent Level of vessel HP Separator and the Actuator DesiredPosition of VLV-100 as OP.

70

5.3.6. The Range for the PV is now easy to define (0% to 100%), the action shouldbe Direct (if PV decreases, OP needs to decrease; if PV increases, OP needs to increase),and first initial guesses for Tuning parameters can be 2 and 10 for Kc and Ti, respectively.Normally, SP for liquid level in vessels is 50%.

5.3.7. Open the Face Plate of the LIC-100.

The IP and LP separators are different than the HP Separator. They are separating 3phases and consequently, they contain 2 inter-phase levels (Liquid 1 - Liquid 2 and Liquid 1 -Vapor). For a proper control of the 3 phases, 2 level controllers are needed for each separator.

71

The same PID controller type that has been used for the HP Separator can be used tocontrol the level of the light liquid in the IP and LP Separators. Because the water productis not very important for the plant profitability, an accurate and expensive control is notneeded and a simple On-Off controller is going to be implemented to control the water levelin the boot of the separator.

5.3.8. Install a PID controller and select the Liquid Percent Level in IP Separator as PVand Actuator Desired Position (ADP) of VLV-101 as OP.

5.3.9. Use the guidelines in point 3.8 to define it.

5.3.10. Open its Face Plate.

5.3.11. Repeat steps 3.10 to 3.12 for the LP Separator, using the ADP of VLV-102 asthe OP.

An On-Off for Water level

When we had a look to the nozzles position we noted that the light liquid nozzle wasat 33% of vessel height and that the heavy liquid nozzle was at 0% of vessel height. So, itis important to avoid water to accumulate in the vessel to a level above this 33% to preventwater flowing out of the vessel through the light liquid nozzle. When defining the On-Offcontroller parameters this needs to be taken into account.

An On-Off controller is implemented using the functionality of the Digital Point logicaloperator: that is located in the Object Palette, just above the PID controller.

5.3.12. Install one Digital Point logical operator close to the IP Separator and use theHvyLiquid Percent Level variable as Process Variable. For the OP, like we have been do-ing up to now, it is necessary to select the Actuator Desired Position of VLV-103. However,for the Digital Point controller it is necessary to select the Digital Actuator Desired Position.

72

5.3.13. Define its Parameters according to the picture below:

73

5.3.14. Install an identical On-Off Controller for the LP Separator.

5.3.15. Switch to Dynamics mode and start the Integrator for some minutes. Observehow the control strategy reacts, how the controlled variables (PVs) stay at the desired SP(or not) and how the manipulated variables (OP) change to drive the PVs to the SPs.

Working in this way, we can only see what is happening at every moment, but we are notstoring any historical data of what has been happening before. In order to analyse historical

74

behavior, we need a tool to visualize the evolution of the variables and to store a number ofhistorical points. This can be achieved using the Strip charts.

Strip Charts can be created by the user selecting the variables of interest or can be au-tomatically created by the simulator. In either of the cases, they can be later modified tochange the variables displayed.

5.3.15. Double-click on the HP Separator to open its property view and move to theStrip Chart page of the Dynamics tab. Select Small Dynamic State in the drop-down menu.Remove part of the variables to leave only:

5.3.16. Press the Create Stripchart... button. A new window opens with the just createdgraphical interface:

75

5.3.17. Open the Databook (Ctrl + D) and see in its Variables tab that the three HPSeparator variables have been automatically added. If you move to its Strip Charts tab,then you will see that the three variables are selected for the only existing Strip Chart, HPSeparator-DL1.

5.3.18. Still in the Strip Charts tab, press the Add button to create a new one. Changeits default name to IP Separator.

5.3.19. Go back to the Variables tab and press the Insert button to find the relevantvariables for the IP Separator:

• Liquid Percent Level.

• HvyLiquid Percent Level.

• Vessel Pressure.

• Vessel Temperature.

5.3.20. Go back to the Strip Charts tab, highlight the IP Separator strip chart and acti-vate the checkboxes of the four variables mentioned above.

5.3.21. In the View group, press the Strip Chart... button.

5.3.22. For the time being, close the Databook window.

76

5.4 Completing the model

Up to now, we have been taking care of the modeling of the liquids in the flowsheet. Allvapor streams were left as boundary streams, without too much modeling. The simulationcase is now going to be completed by adding the vapor compressors and related equipment.The final setup of the processing plant is the one shown in the picture below:

5.3.23. Open the case you saved at the end of point 3.31 in Exercise 3, Inlet Plant-Exercise3. Your case should be in Steady State Mode.

5.3.24. Switch the Color Scheme back to the original Default color scheme.

5.3.25. The LP vapor streams are going to be compressed up to the IP vapors pressureand mixed together. To do so, install a Mixer downstream of IP Vap stream and set it toEqualize All pressures. Rename the stream at the mixer outlet: To Aftercooler.

5.3.26. Install a compressor K-100 with LP Vap as input and a new stream called IP Vap2 as outlet. Connect this new stream as the second input to the mixer created in point 4.2.The flowsheet should solve completely.

5.3.27. Add a cooler (IP Aftercooler) to decrease the temperature of To Aftercoolerstream down to 38 AoC. There is a 50 kPa pressure drop in the cooler.

5.3.28. Add a cooler (IP Aftercooler) to decrease the temperature of To Aftercoolerstream down to 38 AoC. There is a 50 kPa pressure drop in the cooler.

5.3.29. Install a separator (IP Scrubber) to separate the liquids formed during cooling.It should use as feed the outlet of the recycle block.

5.3.30. Install another compressor (K-101) to increase the pressure of the separator va-pors and a cooler to decrease its temperature down to 38 AoC. A 50 kPa pressure drop canbe assumed in the cooler.

77

5.3.31. Install a Mixer, with Equalize All option, that should mix together these com-pressed and cooled down vapors with the ones that were coming from the HP Separator.Name Gas Export the resulting mixed stream.

5.3.32. Due to the cooling, stream Gas Export is not only vapor. Another separator isneeded (TEG Scrubber) to eliminate the liquids and to have the vapors ready for dehydra-tion once warmed up to 32 AoC in heater H-202 (50 kPa pressure drop).

5.3.33. Use a Mixer (Equalize All) to mix the liquids of the TEG Scrubber with theliquids of the HP Separator. A valve will be needed for the TEG Scrubber liquids line.

5.3.34. Use a Mixer (Equalize All) to mix the liquids of the IP Scrubber with the liquidsof the IP Separator before the Condensate Heater. A valve will be necessary for the IPScrubber liquids line.

78

5.3.35. Now, this part of the flowsheet should look very close to the following diagram.

79