Accounting for changes in investment flows in a soft-linked hybrid model
1 like683 views
The IntERACT model was developed to identify cost-efficient policies to further Denmark's transition to a low-carbon economy by 2050 using a hybrid approach. It soft-links a technology-explicit bottom-up TIMES-DK energy system model with a top-down general equilibrium economic model. The models are iteratively solved to account for feedback between energy prices and investments. Energy service demands and fuel costs from TIMES-DK are transferred to the CGE model, while adjusted demands are fed back. This allows investment flows resulting from changes in energy prices to be considered. The presentation provides an overview of the model setup, linking methodology, and progress implementing the automated iterative linking between models.
Accounting for changes in investment flows in a soft-linked hybrid model
- 1. Accounting for changes in investment flows in a soft- linked hybrid model Maurizio Gargiulo E4SMA Srl, Turin, Italy. Environmental Research Institute, University College Cork, Ireland Kristoffer Steen Andersen Ph.D student DTU Management Engineering
- 2. 2 TIMES-DK and IntERACT developers
- 3. Agenda • What is IntERACT • Modelling approach • Model setup and Linking approach • Cockpit • Summary 3
- 4. What is IntERACT? • Purpose-built hybrid model for identifying cost efficient policies to further the Danish transition to a low carbon emission economy by 2050 Scenario 2012 2020 2050 Short term Medium term Long term Alternative (New policy)Baseline 4 www.ens.dk/interact
- 5. IntERACT: Hybrid model capturing the Danish economy and energy system Technology explicitness Behavioral realism Top-down approach Bottom-up approach General equilibrium feedback IntERACT Note: Adapted from Jaccard (2009) : TIMES-DK model : CGE model 5
- 6. Model setup TIMES-DK energy system model General Equilibrium Model - 20 sectors - Utility maximisation - Profit maximisation Change in demand for energy services on a sector level • Price of energy services • Price of intermediates such as electricity and district heating • Energy subsidies and taxes IntERACTMODEL 6 Technology explicit Behavioural realism General equilibrium feedback
- 7. Energy service as a crux for linking K L E i B? M m1 m2 mi… s=0 s=* s=* s=* s=* Coal Oil DH..Gas Biomass… s=0 DHgasoilcoal … Biom. Energy service Bottom-up fuel mix * Elasticities haven been estimated in the spring of 2015, find more info at: www.ens.dk/interact 7 • This is implemented in the CGE using price-wedge and as Leontief-shares. • Agents maximise profit and utility using the costs of the energy service (relative prices as usual). • The assumption is that firms and households make economic decision based on the (relative) prices of energy services. • TIMES-DK determines the price of energy services and fuel mix (technology decisions happens in TIMES-DK).
- 8. Accounting investment flows • This soft-linking approach transfer energy (service) prices and fuel uses from TIMES to the CGE-model. • Using the properties of the standard zero profit condition in the CGE-model the change in investment flow (capital demand) is calculated residually based on the change in energy service price and the change in fuel input. • This suggest one possible approach for taking investment flows from bottom-up models into account. • This approach is implicitly based on the assumption that only capital input and fuels input changes between different TIMES-DK scenarios. • However if other inputs (labour and materials) also changes significantly between the scenarios then the calculated change in investment flow (capital demand) may suffer from a bias of some order of magnitude. 8
- 9. Baseline (calibration) IntERACT cockpit (MS Excel) Defining baseline and alternative scenarios Top-down assumptions: Growth assumptions, elasticities, macro closure. Bottom-up assumptions: Energy service demand, Fuel cost, Technology catalog CGE productivity indicies 2015-2050 Alternative scenarios (iteration) TIMES -> CGE Electricity and district heat prices, fuel mix, subsidies and taxes (%TIMESscenario%.gdx) CGE->TIMES Adjusted demand projection (%ZZ-CGE_Linking%.dd) TIMES -> CGE Electricity and district heat production and prices Fuel use supply sector (%TIMESbaseline%.gdx) TIMES-DK Baseline scenario 2010-2050 CGE alternative Alternative scenarios 2010-2050 TIMES-DK Alternative scenario 2010-2050 CGE reference Baseline scenario (2010-2050) Report (2010-2050) 9
- 10. Model setup 10 Baseline scenario • Common data (e.g. Economic projection from Danish Ministry of Finance) • The CGE model is calibrated using Economic projection from the Danish Ministry of Finance and Information, and TIMES-DK stand alone optimisation to set up some prices and fuel mix. • CGE productivity indices are calculated so that the CGE-model match the Energy system in the TIMES-DK baseline. Alternative scenarios - Models are soft-linked and solved in an interative process (5 ys steps up to 2050) - The main tests have been done for power sector only model (3-5 iterations to solve the combined models).
- 11. IntERACT: Cockpit The model is run in its own MS Excel user-interface 11
- 12. Summary • IntERACT model developed a generic and rather efficient methodology for linking a TIMES-model to CGE-model • This methodology offers a way of taking the macro economic impact of changes in investment flow from TIMES model into account • Currently the automated iterative linking is functioning for power sector; exchanging information related to prices, production and demand for electricity and district heat. • Work is currently underway testing the iterative linking and expanding it to the other parts of the energy system. 12
- 13. Thank you for your attention For more information, feed-back or question please contact: Kristoffer Steen Andersen Ph.D student DTU Management Engineering +45 6060 3159 krisan@dtu.dk Maurizio Gargiulo gargiulo.maurizio@gmail.com E4SMA Srl, Turin, Italy. www.e4sma.com, www.linkedin.com/company/e4sma, https://twitter.com/E4SMAsrl Environmental Research Institute, University College Cork, Ireland 13