calcFeDemandONONSPEC: Calculate historic and projected other non-specified energy...
In pik-piam/mrremind: MadRat REMIND Input Data Package

calcFeDemandONONSPEC

R Documentation

Calculate historic and projected other non-specified energy demand

Description

Project the IEA flow ONONSPEC into the future. As we have no idea, where this energy demand comes from, we use a very generic methos to project it into the future: We assume an asymptotic model. It starts from the level x_\text{EOD} at the end of data (EOD) with EOD slope \dot{x}_\text{EOD} and approaches the fraction \varepsilon of this slope within \Delta t. Both x_\text{EOD} and \dot{x}_\text{EOD} are determined through a linear regression of the last n time steps with IEA data.

x(t) = x_\text{EOD} + \dfrac{\dot{x}_\text{EOD}}{c} \cdot [1 - \exp (-c \cdot (t - t_\text{EOD}))]

with the decay rate c = -\dfrac{\ln \varepsilon}{\Delta t}
\Delta t is differentiated by scenarios thus approaching a low, med and high long-term level. To assure that scenarios don't differ at the end of history (EOH), time steps between EOD and EOH are projected with med value of \Delta t.

Usage

calcFeDemandONONSPEC(scenario, eoh)

Arguments

`scenario`	character vector of remind demand scenarios
`eoh`	numeric, end of history: last time step without scenario differentiation

Details

Each scenario s has a differentiated \Delta t_s. For \dot{x}_\text{EOD} > 0, the high (low) scenario assumes a longer (shorter) time span and for \dot{x}_\text{EOD} < 0 vice versa to reach a higher (lower) long-term value. We want to make sure that until the end of history (EOH), all scenarios are still identical. So we take the med parameterisation until EOH. Afterwards, we adjust the model such that we start with EOH level and slope and still reach the target slope \varepsilon \cdot \dot{x}_\text{EOD} until t_\text{EOD} + \Delta t_s:

x_s(t) = x_\text{EOH} + \dfrac{\dot{x}_\text{EOH}}{c_s} \cdot [1 - \exp (-c_s \cdot (t - t_\text{EOH}))]

with the decay rate c_s = -\dfrac{\ln \varepsilon } {\Delta t_s - (t_\text{EOH} - t_\text{EOD})} \cdot \left(1 - \dfrac{t_\text{EOH} - t_\text{EOD}} {\Delta t_\text{med}} \right)