Energy conversion - REMIND-MAgPIE
| Corresponding documentation | |
|---|---|
| Previous versions | |
| Model information | |
| Model link | |
| Institution | Potsdam Institut für Klimafolgenforschung (PIK), Germany, https://www.pik-potsdam.de. |
| Solution concept | General equilibrium (closed economy)MAgPIE: partial equilibrium model of the agricultural sector; |
| Solution method | OptimizationMAgPIE: cost minimization; |
| Anticipation | |
Conversion
The core part of the energy system is the conversion of primary energy into secondary energy carriers via specific energy conversion technologies. Around fifty different energy conversion technologies are represented in REMIND. In general, technologies providing a certain secondary energy type compete linearly against each other, i.e. technology choice follows cost optimization based on investment costs, fixed and variable operation and maintenance costs, fuel costs, emission costs, efficiencies, lifetimes, and learning rates. REMIND assumes full substitutability between different technologies producing one energy type.
The secondary energy carriers included in REMIND are:
- Electricity ? used for stationary sector and light duty vehicles.
- District heat and local renewable heat ? used for the stationary sector.
- Hydrogen ? used for stationary sector and light duty vehicles.
- Liquids ? used for the stationary sector and transport sector.
- Solid fuels ? used for the stationary sector.
- Gases ? used for the stationary sector.
REMIND represents each technology by a number of characteristic parameters
- Specific overnight investment costs that are constant for most technologies and decrease due to learning-by-doing for some relatively new technologies (see below)
- Cost markups due to financing costs over the construction time
- Fixed yearly operating and maintenance costs in percent of investment costs
- Variable operating costs (per unit of output, excluding fuel costs).
- Conversion efficiency from input to output
- Capacity factor (Maximum utilization time per year). This parameter also reflects maintenance periods and other technological limitations that prevent the continuous operation of the technology.
- Technical lifetime of the conversion technology in years.
- If the technology experiences learning-by-doing: initial learn rate, initial cumulative capacity, as well as floor costs that can only asymptotically be approached
REMIND represents all technologies as capacity stocks with full vintage tracking. Since there are no hard constraints on the rate of change in investments, the possibility of investing in different capital stocks provides high flexibility for technological evolution. However, the model includes cost mark-ups for the fast up-scaling of investments into individual technologies; therefore, a more realistic phasing in and of technologies is achieved. The model allows for pre-mature retirement of capacities before the end of their technological life-time (at a maximum rate of 4 %/year), and the lifetimes of capacities differ between various types of technologies. Furthermore, depreciation rates are relatively low in the first half of the lifetime and increase thereafter.
Each region is initialized with a vintage capital stock and conversion efficiencies are calibrated to reflect the input-output relations provided by IEA energy statistics (IEA 2007a; IEA 2007b). The conversion efficiencies for new vintages converge across the regions from the 2005 values to a global constant value in 2050. Furthermore, for some fossil power plants, transformation efficiencies improve exogenously over time. Finally, REMIND adjusts by-production coefficients of combined power-heat technologies (CHP) by region to meet the empirical conditions of the base year.
Only one technology converts secondary energy into secondary energy, which is the production of hydrogen from electricity via electrolysis.
Electricity
Around twenty electricity generation technologies are represented in REMIND, see Table 4, with several low-carbon (CCS) and zero carbon options (nuclear and renewables).
Table 4. Energy Conversion Technologies for Electricity (Note: * indicates that technologies can be combined with CCS).
For variable renewable energies, we implemented two parameterized cost markup functions for storage and long-distance transmission grids - see Section Grid and Infrastructure. To represent the general need for flexibility even in a thermal power system, we included a further flexibility constraint based on Sullivan et al. (2013).
The techno-economic parameters of power technologies used in the model are given in Table 5 for fuel-based technologies and Table 6 for non-biomass renewables. For wind, solar and hydro, capacity factors depend on grades, see Section Non-biomass renewables.
Table 5. Techno-economic characteristics of technologies based on exhaustible energy sources and biomass (Iwasaki 2003; Hamelinck 2004; Bauer 2005; Ansolabehere et al. 2007; Gül et al. 2007; Ragettli 2007; Schulz 2007; Uddin and Barreto 2007; Rubin et al. 2007; Takeshita and Yamaji 2008; Brown et al. 2009; Klimantos et al. 2009; Chen and Rubin 2009).
For abbreviations see Table ?Acronyms and Abbreviations? ; * for joint production processes; § nuclear reactors with thermal efficiency of 33%; # technologies with exogenously improving efficiencies. 2005 values are represented by the lower end of the range. Long-term efficiencies (reached after 2045) are represented by high-end ranges.
Table 6. Techno-economic characteristics of technologies based on non-biomass renewable energy sources (Neij et al. 2003; Nitsch et al. 2004; IEA 2007a; Junginger et al. 2008; Pietzcker et al. 2014).
Heat
REMIND also features a broad range of technologies for the supply of non-electric secondary energy carriers, such solids, liquids, gases, heat and hydrogen.
Table 7a. Conversion Technologies for heat.
Grid and infrastructure
General distribution costs
REMIND represents electricity/gas/hydrogen grids as well as distribution costs for solids and liquids in terms of a linear cost-markups on final energy use.
Variable renewable energy sources
Variable renewable electricity (VRE) sources such as wind and solar PV require storage to guarantee a stable supply of electricity (Pietzcker et al. 2014b). Since the techno-economic parameters applied to CSP include the cost of thermal storage to continue electricity production at nighttime, REMIND assumes that CSP requires only limited additional storage for balancing fluctuations.
The approach used in REMIND follows the idea that storage demands for each VRE type rise with increasing market share. This is because balancing fluctuations becomes ever more challenging with higher penetration[1].
For modeling reasons, there is a ?generalized storage unit?, tailor-made for each VRE. This construct consists of a VRE-specific mix of short- and medium-term storage as well as curtailment. Examples are redox-flow batteries for short-term storage, electrolysis and hydrogen storage for medium-term storage, as well as curtailment to balance seasonal fluctuations. A specific combination of these three real-world storage options is determined in order to match the VRE-specific fluctuation pattern. From this combination of actual storage technologies, we calculate aggregated capital costs and efficiency parameters for the ?generalized storage unit? of a specific VRE.
To calculate the total storage costs and losses at each point in time, the calculated ?generalized storage unit? of a VRE is scaled with this VRE?s scale-factor ?VRE. The capital costs of the generalized storage units decrease through learning-by-doing with a 10% learning rate.
Costs for long-term HVDC transmission are included following a similar logic as storage costs. REMIND assumes that grid requirements increase with market share. Furthermore, since resource potentials for PV (suitable for decentralized installation) are not as localized as those for wind and CSP, REMIND assumes that grid costs for PV are comparatively smaller.
Both storage and grid requirements are partly regionalized: in regions where high demand coincides with high wind (EUR) or solar (USA, ROW, AFR, IND, MEA) incidence, storage requirements are slightly reduced. If a region is small or has homogeneously distributed VRE potentials (EUR, USA, IND, JPN), grid requirements are lower.
For a market share of 20%, marginal integration costs (including storage, curtailment and grid costs) are in a range of 19-25 USD/MWh for wind, 20-35 USD/MWh for PV, and 8-15 USD/MWh for CSP. For more details on the modeling of VRE integration in REMIND, see Pietzcker et al. (2014b).
Carbon capture and Storage
Deployment of carbon capture and storage (CCS) can curb emissions from fossil fuel combustion. In REMIND, CCS technologies exist for generating electricity as well as for the production of liquid fuels, gases, and hydrogen from coal. Moreover, it is possible to combine biomass with CCS to generate net negative emissions. Such bioenergy CCS (BECCS) technologies are available for electricity generation (e.g., biomass integrated gasification combined cycle power plant), biofuels (e.g., biomass liquefaction), hydrogen, and syngas production. The sequestration of captured CO2 is explicitly represented in the model by accounting for transportation and storage costs (Bauer 2005). There are regional constraints on CO2 storage potentials. In total, the global storage potential amounts to around 1000 GtC. It is smaller for EUR with 50 GtC, Japan with 20 GtC, and India with 50 GtC. The yearly injection rate of CO2 is assumed not to exceed 0.5% of total storage capacity due to technical and geological constraints. This creates an upper limit of 5 GtC per year for global CO2 injection.
[1] Current electricity systems already require substantial flexibility due to varying demand. This flexibility allows for the use of low shares of individual VRE (below ~10%) without any adaptations or storage requirements, as seen in many of today?s electricity networks. Furthermore, many regions have some limited potential for (cheap) pumped hydro storage, leading to low storage costs at low market shares of VRE.