Energy process modeling simulation

System Modelling & Simulation of Energy

Systems

References: Hodge, B.K., and Taylor, R.P., 1999, Analysis and Design of Energy Systems, 3rd edition, Prentice-Hall, Inc. Stoecker, W.F., 1989, Design of Thermal Systems, 3rd edition, McGraw-Hill.

Process Modeling and Simulation

� Process modeling today is an integral part of process development and design, operation optimization, process control, business evaluation and decision-making.

� Modeling and simulation including computer simulation / calculation software or program and mathematical representations of physics and chemistry of complex process system have been increasingly used to assist process development and design.

Three general approaches to modeling:

� First principle approach� Functional block approach� Gray box approach� In the first principle approach, the models are

derived based on physical and chemical laws, e.g. mass and energy balances, thermodynamic equilibrium, chemicalreaction kinetics and mass and heat transport phenomena.

The functional block / black box approach

� The functional block is often referred as black box approach. In this approach, the models are derived strictly based on empirical descriptions. In most cases a complete process model building is not entirelybased on a single approach. A process model built by the first principle approach usuallyleaves a few parameters to be validated by experimental or the real plant operation data.

gray box approach

� A process model built by the first principle approach usually leaves a few parameters to be validated by experimental or the real plant operation data.

� On the other hand, the mathematical structure of an empirical process model should be assumed based on an understanding of the physicochemical nature of the process. This combined approach is often called gray box approach.

Two types of models are:

� Steady state model� Dynamic model� A steady state model reflects the process

during steady state operation. Neither energy nor material accumulations with respect to time are considered in the model.

� A dynamic model reflects the time transient response of the process from one steady state to another. Energy and / or material accumulations with respect to time are considered in the model.

Modeling vs. Simulation

� Modeling deals with establishing physically correct quantitative relationships between real systems and models of those real systems

� Simulation deals with implementing the models, usually using the computer, in such a way that the results match those of the real system to a high degree

Degree of validity: the extent to which

the model matches data from the real

system

� Replicatively valid: it matches existing data from the real system

� Predictably valid: it can predict data outside the range of parameters of the original database

� Structurally valid: it truly reflects how the real system operates

� All simulations should be validated using experimental data.

Purpose of Simulation

� Simulations can have a wide range of purposes. They may include

� Predicting off-design performance of existing systems to identify and mitigate possible problems

� Optimizing the efficiency of a system during the design process to decrease energy costs

� Determining how making a modification in one part of an existing system will affect the rest of the system.

� Once validated with experimental data, simulations can save a lot of time and money – they’re a lot cheaper and faster than running experiments!

Classes of Simulations

� Continuous vs. Discrete � Continuous: flow through a system is continuous, like

fluid flow � Discrete: flow is treated as a certain number of

discrete integers, such as number of people � Deterministic vs. Stochastic � Deterministic: input parameters are known and

precisely specified � Stochastic: input parameters are uncertain. They may

be determined randomly or using a probability distribution, for instance.

� Steady State vs. Transient (Dynamic) � Most of our problems will be continuous,

deterministic, and steady state

Developing accurate models and

simulations

� Several areas must be looked at closely to develop an accurate simulation

� Physical bases

� Levels of the component models

� Accuracy

� Validation procedure

Physical bases

� If the component models don’t represent the correct physics, the model will not give accurate results or you will not be able to use the model beyond a very limited range

� How do the individual components act? � Do your mathematical equations accurately predict

performance? � Do you understand how the different components

interact? � Make sure that you include the effects that the system

may have on the component performance.

Levels of Component Model

� The higher the level of the model, the more details are captured. For example, think of a compressor model

� Level 1 might be doing a simple analysis like done as homework problems for an undergraduate thermodynamics class.

� Level 2 might be the model that one developes for more parameters.

Levels of Component Model

� Level 3 might be a detailed transient finite difference computer analysis of the fluid dynamics inside the compressor.

� Higher level models, if done correctly, are more accurate and model the true situation better.

� However, you pay a price with increased computation time and increased personal time to develop them.

Accuracy

� Make sure that you clearly understand the assumptions being made and how they affect the accuracy of the results.

� Choose a simulation level consistent with your desired accuracy.

� You may have no need for a sophisticated finite difference model.

� Use a similar simulation level for each component in your system unless you have a good reason not to.

Accuracy

� Your system accuracy might be dominated by the component modeled with the lowest-level model. In that case, there’s no reason to use a higher-level model for other components.

� Performing a sensitivity analysis may help us determine how good the model of a certain component should be.

� If “y” is the desired output and “x” the result of an individual component model, vary “x” and see how much ”y” changes. If changing “x” has little effect on “y”, then that component doesn’t need a very sophisticated model.

Validation

� This includes two steps: validating the individual component models and the entire system simulation.

� Make sure that your simulation can reproduce existing experimental data.

� Be careful about running your simulation for parameters outside the range of validation. At times you may need to, but realize that you’re increasing the uncertainty of your results.

Who Should do Simulations

� Computer scientists and computer engineers may be

able to assist in transforming your model into a

computer simulation

� However, they will not understand the physics that

you’re trying to model, how components interact, the

required accuracy, what assumptions are OK, etc.

� Therefore, it is important that system simulations be

developed by the people who are experts in that

particular area of engineering.

Systems of Simultaneous Equations

� Many simulations require solving systems of simultaneous equations.

� Finite difference methods� Using EES for a refrigerator model, the compressor

performance depends on the evaporator, which depends on the expansion valve, which depends on the condenser, which depends on the compressor –they’re all linked together. Two methods of simulation include Successive Substitution and Newton -Raphson

ENERGY CONVERSION PLANT

ENERGY CONVERSION PLANT

Dynamic Life Cycle Assessment of biogas

production from micro-algae

Because of their high production yield, micro-algae have been pointed as an interesting biofuel. A relevant mean to upgrade the energy value of micro-algae with optimal performances is the anaerobic digestion of the algae. It enables achievement of environmental benefits and production of energy from renewable resources. However such processes only exist at lab-scale.

Continued...

� In order to assess and optimize its performances and environmental impacts, one has to simulate its behaviour through dynamical models.

� In broad outline the two major compartments of the system (micro-algae culture and anaerobic digestion process) are linked by internal flows (micro-algae, digestates…) and receive external flows (light, cosubstrates…). As a consequence, overall behaviour is determined by the interaction of several time-dependent processes.

Biogas production from micro-algae

� We integrate dynamic

system modeling of

micro-algae growth and

anaerobic digestion of

biomass in the LCA in

order to obtain dynamic

flows.

Continued...

� A pertinent Life Cycle Inventory can not be achieved without taking into account the dynamic of several processes; some economic flows are determined according to the temporal evolution of processes.

� Consequently, we integrate dynamic system modeling of micro-algae growth and anaerobic digestion of biomass in the LCA in order to obtain dynamic flows.

� This approach allows us to obtain dynamic data for the Life Cycle Inventory. This is a preliminary step to more accurate impact assessment.

Wind Energy Conversion System

The Modelling Process

Real world

problem

Mathematical

problem

Mathematical

solutionInterpretation

1

2

3

4

Model Application Areas

Process design

Process control and diagnosis

Troubleshooting

Process safety

Operator training

Environmental impact assessment