Turbinmodellering – En jämförelse av polygonmodellering och komponentmodellering
(2018) MVKM01 20181Heat Transfer
- Abstract
- Abstract
In order to be able to calculate how an energy system should be controlled to achieve minimal overall cost in advance,energyproducers often use software for economic operational optimization. In order for the software to make good optimizations, the model of the energy system needs to be realistic. It is also important for the user that optimizations do not take too long and that it is easy to make changes to the model. In this project, a method was developed to build a turbine model in an already existing energy system model using so-called polygon modeling. The idea was to be able to turbine model to find out which one had the most desirable properties.
The polygon model was built up by analyzing statistic data from the real... (More) - Abstract
In order to be able to calculate how an energy system should be controlled to achieve minimal overall cost in advance,energyproducers often use software for economic operational optimization. In order for the software to make good optimizations, the model of the energy system needs to be realistic. It is also important for the user that optimizations do not take too long and that it is easy to make changes to the model. In this project, a method was developed to build a turbine model in an already existing energy system model using so-called polygon modeling. The idea was to be able to turbine model to find out which one had the most desirable properties.
The polygon model was built up by analyzing statistic data from the real turbine. With two years of collected measurement data, a relation was setup to determine an equation that will predict future loadcases.The developed equation was implemented along with additional constraints in the polygon model and comparison with the existing model was made.
In this study, Energy Opticon’s Energy Optima3 software was used to build the polygon model in an existing client’s energy system. It was found that the built-in polygon model had 23.8 percentage points lower overall error, 21.2 percentage points lower average error and a reduced standard deviation of 3.7 MW than the existing turbine model. An improvement in the existing model resulted in the polygon model having 2.7 percentage points lower overall error, 4.6 percentage points lower average error and a reduced standard deviation of 0.54 MW than the modified model. Furthermore, it was found that the optimization of the polygon model was slightly slower than the optimization time for the existing model. Both turbine models were considered as easy to adjust, on the other hand, the existing model was considered to be somewhat more easily understood. An advantage of the polygon model was that no turbine specifications were needed to build the model, instead, collected measured data was used to statistically predict how the turbine should behave at given input.
The biggest reasons for the smaller error of the polygon model were the simplifications of the existing model regarding drainage in the turbine and the minimum load of the existing turbine components. The minimum load of the turbine components prevented the model from achieving load cases that the real turbine was able to achieve.
In summary, polygon modeling is an interesting substitute for today’s component modeling, not least for those customers advocating high precision rather than making the model easy to understand. In particular, polygon modeling is beneficial for those energy producers who lack complete specifications for the turbine in their energy system. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/8952201
- author
- Carl-Johan, Kokacka and Olof, Levin
- supervisor
-
- Marcus Thern LU
- organization
- alternative title
- Turbine Modeling - A Comparison of Polygon Modeling and Component Modeling
- course
- MVKM01 20181
- year
- 2018
- type
- H2 - Master's Degree (Two Years)
- subject
- report number
- LUTMDN/TMHP-18/5411-SE
- ISSN
- 0282-1990
- language
- Swedish
- id
- 8952201
- date added to LUP
- 2018-06-21 14:08:32
- date last changed
- 2018-06-21 14:08:32
@misc{8952201,
abstract = {{Abstract
In order to be able to calculate how an energy system should be controlled to achieve minimal overall cost in advance,energyproducers often use software for economic operational optimization. In order for the software to make good optimizations, the model of the energy system needs to be realistic. It is also important for the user that optimizations do not take too long and that it is easy to make changes to the model. In this project, a method was developed to build a turbine model in an already existing energy system model using so-called polygon modeling. The idea was to be able to turbine model to find out which one had the most desirable properties.
The polygon model was built up by analyzing statistic data from the real turbine. With two years of collected measurement data, a relation was setup to determine an equation that will predict future loadcases.The developed equation was implemented along with additional constraints in the polygon model and comparison with the existing model was made.
In this study, Energy Opticon’s Energy Optima3 software was used to build the polygon model in an existing client’s energy system. It was found that the built-in polygon model had 23.8 percentage points lower overall error, 21.2 percentage points lower average error and a reduced standard deviation of 3.7 MW than the existing turbine model. An improvement in the existing model resulted in the polygon model having 2.7 percentage points lower overall error, 4.6 percentage points lower average error and a reduced standard deviation of 0.54 MW than the modified model. Furthermore, it was found that the optimization of the polygon model was slightly slower than the optimization time for the existing model. Both turbine models were considered as easy to adjust, on the other hand, the existing model was considered to be somewhat more easily understood. An advantage of the polygon model was that no turbine specifications were needed to build the model, instead, collected measured data was used to statistically predict how the turbine should behave at given input.
The biggest reasons for the smaller error of the polygon model were the simplifications of the existing model regarding drainage in the turbine and the minimum load of the existing turbine components. The minimum load of the turbine components prevented the model from achieving load cases that the real turbine was able to achieve.
In summary, polygon modeling is an interesting substitute for today’s component modeling, not least for those customers advocating high precision rather than making the model easy to understand. In particular, polygon modeling is beneficial for those energy producers who lack complete specifications for the turbine in their energy system.}},
author = {{Carl-Johan, Kokacka and Olof, Levin}},
issn = {{0282-1990}},
language = {{swe}},
note = {{Student Paper}},
title = {{Turbinmodellering – En jämförelse av polygonmodellering och komponentmodellering}},
year = {{2018}},
}