Model design Flashcards
Examples of model expressions:
CAPEX OPEX (P)EROI Emissions - carbon footprint EcoPoints
Expand: CAPEX
Capital expenditures (CAPEX) are a company's major, long-term expenses Examples - buildings, equipment, machinery, and vehicles.
Expand: OPEX
Operating expenses are the costs a company incurs for running its day-to-day operations.
Examples: Wages, rent, taxes
(P)EROI
(Processed) Energy Returned or Invested
Energy Return on Investment (EROI) is a ratio for describing a measure of energy produced in relation to the energy used to create it.
The threshold between energy producer and energy sink is set at 1.
Temporal influences on overall sustainability
Energy demand, intermittent resources, decentralised load balancing
Factors influencing system sustainability
LCA characteristics
Temporal dynamics
Geographic diversity
Complexity
Relative patterns
% of maximum flow available at every interval between 0% and 100%
(cow manure either 100% available from stable or 0% available from the field)
Absolute pattern
Actual value per interval
wind speed, temperature which can be implemented into the calculation of the flow
Model Verification
Verification ensures that the computerised model made correctly
“did I create the model right”
Model Validation
Validation ensures that the model itself functions accurately for the specific purpose
“Did I create the right model?”
Acceptable range of accuracy
The accuracy required of the model for its intended purpose
The cost curve
As the cost increases, so does the confidence in model accuracy
Value curve
Value of the model increases to the user as the confidence of the model increases
The problem entity
The system idea, situation, policy or phenomena to be modelled
The conceptual model
The mathematical, logical, graphical representations of the problem entity
Computerised model
conceptual model implemented on the computer
How is the PE, and Concep and Comput models interlinked
conceptual developed through analysis and modelling
conputerised developed through programming and implementation
Inferenced about PE are obtained through experimentation on computerised models
Operation validity
The models output behaviour has a satisfactory range of accuracy for the intended purpose of the model
Data validity
ensuring necessary data for the model building, evaluation and testing are adequate and correct
Ground truth
A measure of validity..
Information that is known to be real or true.
A model is an approximation of reality - how far is it from the ground truth
What are the primary techniques required for model validation
Face validation, structured walk through, trace
What are the primary techniques used to determine correct programming
Structured walk through, trace `
How can an LCA be used by decision makers `
(RIMC)
Reduce negative impact of new product
Identify what can be improved in existing products
Modifying aspects which can cause more significant issues further in its lifetime
Comparing environmental performance of similar products
Why is the expression of GWP in kgco2eq and not Kgco2?
Measuring the impact of emissions on the environment accounts for all greenhouse gases. Different gases will have different environmental impact - however they are summed up into a single score of GWP 100.
What is EcoPoints?
A single unit measurement of obtained by a combination of factors of impact categories relating to environmental impacts.
3 impact categories - human health, ecosystems and resource depletion.
Handling complexity with modelling
Modular approach
Consciousness of system boundaries
Budget
Time
Components within a single sub-module
Main variables Impact factors Primary flows Direct flows Indirect flows Embodied flows
What kind of information can you find in the primary data base
Physical properties of the materials used in primary flows. for example, material densities, heating values and molar mass.
What data is found within the direct coefficient database
Here will be the values needed for calculating the direct flows - for example, the amount of electricity, heat, diesel required to conduct these processes
What information can be found in the direct IMPACT coefficient data base
Data for calculating direct impact indictors such as the costs (CAPEX and OPEX) efficiency (P)EROI, Emissions (GWP) and environmental impacts (EcoPoints)
What data is found within the indirect data base
specific coefficients relating to the processes needed for the direct material and energy flows - mainly used for calculating impact factors e.g. extraction, refining, transport and use of diesel
What data is found in the embodied database
Impact factors relating to the installations required for the specific sub-module - costs, materials used, construction, energy for constructions and deconstruction (at end of life).
There are two main variables for embodied 1. the size/ power rating of the infrastructure 2. the lifetime of the infrastructure
Questions to ask when verifying a model… Does/is the model
Add to scientific findings/ societal benefits
Provide clear answers
Reviewed and verified by experts
Subjective testing
Testing through common reasoning by the modeler
Can be done using graphical display and review of model behaviour - also comparison to other models
Objective testing
Using some kind of statistical/mathematic test e.g. hypothesis testing or confidence intervals.
How can accuracy be measured
Factual aspects - compared with the ground truth
Conceptual aspects - opinion by experts
What is the link between accuracy and development time
The goal is to retain the highest accuracy in the lowest time required
% of accuracy for fact based comparisons
80%
however primary data accuracy should be in the range of 95%-99%.
Why should you be mindful of the information on EcoPoints
The ReCiPe software is still evolving = possible uncertainties in the indicators
Hypothesis testing
To determine whether the models outputs are in a satisfactory range for its intended purpose - means, distributions, variances and time series are tested under different scenarios.
Confidence intervals
The shorter the lengths of the confidence intervals the more invalid the data is considered - they are the differences between the means, variances and distributions of different simulations and outputs from a set of experimental conditions.
Types of graph for model comparison
Histograms
Box-plots
Behaviour charts
“output behaviour has sufficient accuracy for models intended purpose”
Functional unit
Quantified performance of a product system for use as a reference unit in an LCA study
It is a single comparison unit of an LCA indicated in a measurable unit (MJ, Kg)
Function of SIMAPRO
Science-based tool used to collect, analyse and monitor the sustainability performance data of your company’s product or service. Assessing the environmental impact across all life cycle stages
Limitations of LCA
Defining the system boundary - spatial and temporal resolutions and capital goods analysis.
Not suitable for precautionary approach - catastrophes
Other system boundaries - socioeconomic, financial (subjective)
Specific data difficult to obtain
Time consuming`
Advantages of LCA
System reliability - well accepted Software/database availability Includes multiple impact categories Allows for comparison (also think how decisions makers/companies will use this information for the benefit of themselves and environment)
LCA interpretation should include
Identification of significant issues as proposed in the problem entity
Evaluation of the study - completeness, sensitivity, and consistency
Drawn recommendations and conclusions and limitations
Production chain
Analytical tool used to understand the nature of the production process.
Other types of models
Predictive models - altering the variables to fit different scenarios to extrapolate potential future projections
Optimisation models - altering the data set to find out optimal outcomes for specific values (costs, efficiency).
3 purposes for data collection
building conceptual model
validating the model
experimenting with the validated model
Conceptual model validity
Ensuring the underlying theories and assumptions in relation to the problem entity are correct
The logical, mathematical and graphical assumptions are ‘reasonable’ for the intended purpose of the model
Operational validity
Ensuring the model output behaviour has the accuracy required for the intended purpose of the model.
(this is where subjective and objective testing comes in)
Why can model ‘correctness’ not be fully established
no algorithm exists to test model correctness
Every simulation project is new and unique so must be tailored made and verified throughout
GWP
100 year global warming potential.
Comparable for fossil and renewable energy sources.
Many different GHG with their own potentials and properties.
carbon neutral, carbon releases and fossil fuels.
What is the process of model design
Problem definition
• Conceptual model: Main expressions, Main variables, Equations
• Data collection: Literature, Measurements
• Actual model
• Validation
