willu47
commented
2 months ago There are indeed a number of differences across the two files you shared and the current version in this repo. Taking just the objective function:
osembe_ecemf
minimize cost: sum{r in REGION, t in TECHNOLOGY, y in YEAR}
(((((sum{yy in YEAR: y-yy < OperationalLife[r,t] && y-yy>=0}
NewCapacity[r,t,yy])+ ResidualCapacity[r,t,y])*FixedCost[r,t,y] +
sum{m in MODEperTECHNOLOGY[t], l in TIMESLICE}
RateOfActivity[r,l,t,m,y]*YearSplit[l,y]*VariableCost[r,t,m,y]) / DiscountFactorMid[r,t,y] +
CapitalCost[r,t,y] * NewCapacity[r,t,y] / DiscountFactor[r,t,y] +
DiscountedTechnologyEmissionsPenalty[r,t,y] - DiscountedSalvageValue[r,t,y]) +
sum{s in STORAGE} (CapitalCostStorage[r,s,y] * NewStorageCapacity[r,s,y] / DiscountFactor[r,t,y] -
CapitalCostStorage[r,s,y] * NewStorageCapacity[r,s,y] / DiscountFactor[r,t,y]));osembe
minimize cost: sum{r in REGION, t in TECHNOLOGY, y in YEAR}
((((sum{yy in YEAR: y-yy < OperationalLife[r,t] && y-yy>=0}
NewCapacity[r,t,yy])+ ResidualCapacity[r,t,y]) * FixedCost[r,t,y] +
sum{m in MODEperTECHNOLOGY[t], l in TIMESLICE}
RateOfActivity[r,l,t,m,y] * YearSplit[l,y]*VariableCost[r,t,m,y]) / DiscountFactorMid[r,y] +
CapitalCost[r,t,y] * NewCapacity[r,t,y] * CapitalRecoveryFactor[r,t] * PvAnnuity[r,t] / DiscountFactor[r,y] +
DiscountedTechnologyEmissionsPenalty[r,t,y] - DiscountedSalvageValue[r,t,y]) +
sum{r in REGION, s in STORAGE, y in YEAR} (CapitalCostStorage[r,s,y] * NewStorageCapacity[r,s,y] / DiscountFactorStorage[r,s,y] -
- SalvageValueStorage[r,s,y] / ((1+DiscountRateStorage[r,s])^(max{yy in YEAR} max(yy)-min{yy in YEAR} min(yy)+1)));fast
minimize cost: sum{r in REGION, t in TECHNOLOGY, y in YEAR}
((((sum{yy in YEAR: y-yy < OperationalLife[r,t] && y-yy>=0}
NewCapacity[r,t,yy])+ ResidualCapacity[r,t,y]) * FixedCost[r,t,y] +
sum{m in MODEperTECHNOLOGY[t], l in TIMESLICE}
RateOfActivity[r,l,t,m,y] * YearSplit[l,y] * VariableCost[r,t,m,y]) / DiscountFactorMid[r,y] +
CapitalCost[r,t,y] * NewCapacity[r,t,y] * CapitalRecoveryFactor[r,t] * PvAnnuity[r,t] / DiscountFactor[r,y] +
DiscountedTechnologyEmissionsPenalty[r,t,y] - DiscountedSalvageValue[r,t,y]) +
sum{r in REGION, s in STORAGE, y in YEAR} (CapitalCostStorage[r,s,y] * NewStorageCapacity[r,s,y] / DiscountFactorStorage[r,s,y] -
- SalvageValueStorage[r,s,y] / ((1+DiscountRateStorage[r,s])^(max{yy in YEAR} max(yy)-min{yy in YEAR} min(yy)+1)));To be honest, none of these objective functions are correct. The ECEMF correctly includes a technology-specific version of DiscountFactorMid but omits the correction to storage which is in fast and osembe. osembe and fast include the updated capacity recovery factor * pv annuity, but omit a technology-specific discountfactor.
HauHe
robertodawid
trevorb1
For the European power sector model we us an OSeMOSYS code that contains custom equations for the
ReserveMargin. Therefore, updates to the OSeMOSYS code cannot be directly used.A few days ago I have updated the code using the
osemosys_fastcode from GitHub. However, when running OSeMBE with the updated code the results are notably different from the old code. The scenario I ran is the ECEMF WP1 NetZero scenario which is set-up to reach negative emissions in 2050, i.e., emissions are 0 in a year before 2050. This is done by using an EmissionPenalty. For my use case I have tested the impact that the availability of CCS has on the results. Before 2040 the model can't invest in CCS, only from 2040 onwards it may do so.With the old code the model starts investing into biomass CCS directly in 2040. In 2040 a bit less than 10 time more than in the following years. With the updated code however, the investments into BM CCS 2040 are 10 times higher than with the old code.
I guess this behaviour is not desired. I wonder if this might be caused by something in the updated formulation of discounting and or the issue of considering emissions penalties for negative emissions #99.
I attach the two osemosys versions. osemosys_osembe_ecemf.txt is the old code version. osemosys_osembe.txt the new code version.
osemosys_osembe_ecemf.txt osemosys_osembe.txt