4.1 Energy supply

WITCH includes a comprehensive range of technology options to describe the final use of energy and the generation of electricity. The energy sector is described by a production function that aggregates different factors at various levels and with associated elasticities of substitution \(\rho\). The main distinction is among electric generation and non-electric consumption of energy.

The key technological-economic features represented are: yearly utilisation factors, fuel efficiencies, investment, and operation and maintenance costs, capital depreciation.

The following graph shows the fundamental CES nest structure of the economy and energy system. The number below a node represents the elasticity of substitution at a given nested CES.

CES Production Function of WITCH

CES Production Function of WITCH

4.1.1 Electric energy technology sectors

Electricity is generated by a series of traditional fossil fuel-based technologies and carbon-free options. Fossil fuel-based technologies include natural gas combined cycle (NGCC), fuel oil and pulverised coal (PC) power plants. Coal-based electricity can also be generated using integrated gasification combined cycle (IGCC) production with carbon capture and storage (CCS). Low carbon technologies include hydroelectric and nuclear power, renewable sources such as wind turbines and photovoltaic panels and CSP (wind and solar), and two carbon-free backstop technologies, which represent a basket of technological options far from commercialization.

The cost of electricity generation is endogenous and combines capital costs, O&M expenditure, and the costs for fuels. For nuclear power, waste management costs are also modelled, but no exogenous constraint is assumed.

The following table provides a full list of the CES nodes including all energy options:

Abbreviation Name Abbreviation Name
KL Capital-labor aggregate ELNUKE&BACK Electricity generated with nuclear and backstop
K Capital invested in the production of the final good ELBACK Electricity generated with backstop
L Labor ELNUKE Electricity generated with nuclear
ES Energy services ELW&S Wind turbines and photovoltaic panels
RDEN Energy R&D capital NEL Nonelectric energy
EN Energy TradBiom Traditional Biomass
EL Electric energy COALnel Coal for nonelectric energy
ELYHYDRO Electricity generated with hydroelectric OGB Oil, backstop, gas, and biofuel
EL2 Electricity generation GASnel Gas for nonelectric energy
ELFF Fossil fuel electricity OIL&BACK Oil and backstop for nonelectric energy
ELCOALBIO Electricity generated with Coal and Biomass BACKnel Backstop for nonelectric energy
EPC Electricity generated with pulverized coal OILnel Oil for nonelectric energy
ELIGCC Electricity generated with integrated gasification combined cycle coal plus carbon capture and storage Biofuels Traditional and advanced biofuels
ELOIL Electricity generated with oil Trad Bio Traditional biofuels
ELGAS Electricity generated with gas ELWIND Electricity generated with Wind Energy
ELGASTR Electricity generated with Gas turbines ELPV Electricity generated with Photovoltaics
ELGASCCS Electricity generated with Gas with CCS ELCSP Electricity generated with Concentrated Solar Power
WINDON Electricity generated with Onshore Wind WINDOFF Electricity generated with Offshore Wind
ELPC Electricity generated with Pulverised Coal ELCIGCC Electricity generated with Coal IGCC plus CCS
ELPB Electricity generated with biomass ELBIGCC Electricity generated with Biomass with CCS

4.1.2 Capital accumulation for electric energy technology sectors

The capital of some electricity production technologies accumulates as follows: \[ K_j(t+1,n) = K_j(t,n)\left(1-\delta_j(t+1,n)\right)^{\Delta_t} + \Delta_t \times \frac{I_j(t,n)}{SC_j(t,n)}, \forall j\in \mathcal{J}_{inv}, \] where \(\mathcal{J}_{inv}\) is the set of electric energy technology sectors in which one can invest. For less mature technologies, endogenous technical change is modelled: On the one hand, the investment cost \(SC\) for solar and wind energy is decreasing with the world cumulative installed capacity by means of Learning-by-Doing. On the other hand, the investment costs \(SC\) for backstop technologies are computed using a Two Factor Learning Curve: costs reductions are driven by both, growth in cumulated installed capacity and accumulation of R&D knowledge.

The aforementioned methodology is shown in the Research and Development section.

Given that we use a standard exponential depreciation rule, we calibrate the depreciation rate \(\delta_j(t,n)\) based on a finite lifetime of the power plant with a linear depreciation rate of 1% per year until the end of the life time and full depreciation thereafter. Based on realistic plant specific life times, the exponential depreciation rate is computed equalizing the integral of both depreciation schedules, in order to obtain the equivalent potential output from the capacity.

4.1.3 Production in electric sector

The structure of WITCH in the figure above is represented by the following CES production functions. As above, the value shares \(\alpha\), efficiency parameters \(\phi\), and substitution parameter \(\rho\) are obtained from the Static Calibration of the model. The tree for the power sector is defined by the three following equations combining electricity from fossil fuels \(ELFF\), nuclear and the electrical backstop \(EL_nucback\), and intermittent renewable \(EL_intren\) with hydro power \(EL_hydro\) to electricity \(EL(t,n)\):

\[ EL(t,n) = EL2(t,n) + \alpha_{el_{hydro}}(n) EL_{hydro}(t,n) \]

\[ EL2(t,n) = \phi_{el2}(n) \left( \alpha_{el2}(n) ELFF(t,n)^{\rho_{el2}(n)} + \beta_{el2}(n) EL_{nucback}(t,n)^{\rho_{el2}(n)} + \gamma_{el2}(n) EL_{intren}(t,n)^{\rho_{el2}(n)} \right)^{\frac{1}{\rho_{el2}}(n)} \]

\[ ELFF(t,n) = \phi_{elff}(n) \left( \alpha_{elff}(n) EL_{coalwbio}(t,n)^{\rho_{elff}(n)} + \beta_{elff}(n) EL_{oil}(t,n)^{\rho_{elff}(n)} + \gamma_{elff}(n) EL_{gas}(t,n)^{\rho_{elff}(n)} \right)^{\frac{1}{\rho_{elff}}(n)} \]

4.1.4 Additional constraints in the electric sector

In the electric sector, the production capacity is bounded by the capital stock times the capacity factor \(\mu\) (mu):

\[ EL_j(t,n) \le \mu_j(t,n) \times KEL_j(t,n) \]

For sectors that have are fuel fed, the production capacity, \(Q_j\), is equal to the energy consumed times the overall efficiency rate of the sector \(\xi\):

\[ Q_j(t,n) = \sum_f \xi_{j,f}(t,n) \times Q_{j,f}(t,n) \]

For the backstop technology in the electric sector (advanced nuclear), while Research and Development brings down the future costs (see above), the penetration is limited to 5% of total energy demand in the respective sector per period: \(EL_{elback}\), is bounded by the coefficient lim_elback = 5%.

\[ EL_{elback}(t+1,n) - EL_{elback}(t,n) \le lim\_elback \times \left( EL(t,n) - EL_{elback}(t,n) \right) \]

4.1.5 Non-electric energy sectors

Energy consumption in the non-electric sector is based on traditional fuels (traditional biomass, oil, gas and coal) and bio-fuels. This sector comprises transportation, industrial, and residential and commercial energy use.

The road transport sector is separated out from the CES structure and is described in detail in the section Road Transport (mod_road_transport). The stationary sector comprising industrial, commercial, and residential use of fuels is defined through a CES nest. It combines non-electric energy use from coal \(NEL_coal\), traditional biomass in developing countries \((NEL_trbiomass)\), and an aggregate of Oil&Gas (and traditional biofuels) use \((NEL_log)\). This CES nest is thus defined by the following two equations, where the CES parameters are as before calibrated in the Static Calibration:

\[ NEL(t,n) = \alpha_{nel_{coal}} NEL_{coal}(t,n) + NEL_{log}(t,n) + \alpha_{nel_{trbiomass}} NEL_{trbiomass}(t,n) \]

The \((NEL_log)\) aggregate combines Oil, Gas, and traditional biofuel which are rather easy substitutable in a separate CES nest:

\[ NEL_{log}(t,n) = \phi_{nelog}(n) \left( \alpha_{nelog}(n) NEL_{oilback}(t,n)^{\rho_{nelog}(n)} + \beta_{nelog}(n) NEL_{gas}(t,n)^{\rho_{nelog}(n)} + \gamma_{nelog}(n) NEL_{trbiofuel}(t,n)^{\rho_{nelog}(n)} \right)^{\frac{1}{\rho_{nelog}}(n)} \]

4.1.6 Primary Energy Supply of Fuels

The total primary energy supply of the different fuels is the sum of the consumption from all the sectors:

\[ Q_f(t,n) = \sum_j Q_{j,f}(t,n) \]

The consumption of fuel \(f\), \(Q_f\), equals the total level of extraction by balancing domestic extraction and fuel imports \(X_{f}\):

\[ \sum_n Q_f(t,n) = \sum_n X_f(t,n) \]

If the country is a net exporter, \(X_f\) is negative. The net cost of the different fuels is equal to the cost of extracted fuels consumed fuels minus the cost of the (net-)imported fuels at the world market price \(p_f(t)\)

\[ C_f(t,n) = MC_f(t,n) \times Q_f(t,n) - p_f(t) \times X_f(t,n) \]

The price of fossil fuels and exhaustible resources (\(f\), oil, gas, coal and uranium) is determined by the marginal cost of extraction, \(MC_f\), which in turn depends on current and cumulative extraction. A regional mark-up is added to mimic different regional costs including transportation costs.