Skip to content

DIETER coupling (markup and curtailment)
Compare
Choose a tag to compare
@cchrisgong cchrisgong released this 22 Jul 09:30
v1.0
a4ea098

This is the first stage of coupling between IAM REMIND and energy system model DIETER.
This version corresponds to v1.0 of DIETER coupling model (https://github.com/cchrisgong/dieter-coupling-remind/releases/tag/v1.0).
DIETER repository should be cloned under remind-coupling-dieter/scripts/iterative/ as a submodule.
After cloning remind-coupling-dieter, do "git submodule add https://github.com/cchrisgong/dieter-coupling-remind.git" under remind-coupling-dieter/scripts/iterative

Switch on DTcoup switches (one main switch, one operational switch to turn on the equations after some REMIND iterations) to run the coupled model. Run python DIETER_rewrite.py inside the repository to create parallel models for each year before running the model.

Features including:

  • REMIND informs DIETER of total (usable) power demand, and technical specs: costs (CAPEX, OPEX, fuel cost, CO2 costs), lifetime of plants, efficiency etc
  • DIETER's capacities are bound from below by REMIND's pre-investment capacities (standing capacity before investment in one REMIND period), then make a year-long investment and hourly dispatch. The resulting capacity factors, market values, curtailment ratio are passed back to REMIND
  • Capacity constraint: DIETER informs REMIND the peak hourly residual load demand, REMIND must invest enough dispatchable technologies to meet this peak residual demand
  • Market values are reflected by markups (difference between market value and electricity price), which are imposed as a tax/subsidy in REMIND. Market values in REMIND = electricity price in REMIND + markup in DIETER
  • To ensure better stability, capacity factors are averaged over two iterations.
  • A prefactor to markup is used to ensure the optimization step in investment carried out in REMIND in current iteration matches the current iteration's endogenous market value (decided by current iteration generation share)
  • DIETER's curtailment ratio overwrites that in REMIND for coupled region if coupling is on

Limitations:

  • No ramping costs are included in DIETER, wind offshore is turned off
  • No storage, H2 or DSM are yet implemented. Operational reserves are excluded for simplicity
  • Fuel price from REMIND (which are marginals of primary fuel balance equations) can be volatile, so currently this is fixed to REMIND's first iteration result when passed to DIETER, then smoothed over 3 REMIND periods (usually 15 years). REMIND budget is handled similarly, though no intertemporal smoothening
  • When capacity lower bound in DIETER is slightly less than the REMIND pre-investment capacities, the run is more stable. This is due to the fact that the two models differ in electricity price. This results in DIETER favoring wind in later years than REMIND (see attached graph on generation mix in 2045 for Germany)
  • Certain power generation technologies existing in REMIND are not implemented in DIETER, such as diesel oil turbine
Assets 3
Loading