Commodity market price

[xiao.li8@mcgill.ca]

Hi,

Which attribute can I use to see commodity market price (e.g., electricity)? Do I need to manually calculate it by using other attributes?

Best,
Xiao

[Antti-L]

See the documentation, Part I, section 3.2.4 Marginal value pricing :

"We have seen in the preceding subsections that the TIMES equilibrium occurs at the intersection of the inverse supply and inverse demand curves. It follows that the equilibrium prices are equal to the marginal system values of the various commodities. From a different angle, the duality theory of Linear Programming (chapter 14) indicates that for each constraint of the TIMES linear program there is a dual variable. This dual variable (when an optimal solution is reached) is also called the constraint’s shadow price, and is equal to the marginal change of the objective function per unit increase of the constraint’s right-hand-side. For instance, the shadow price of the balance constraint of a commodity (whether it be an energy form, material, a service demand, or an emission) represents the competitive market price of the commodity."

Based on that, I would say that the shadow price of the balance constraint of a commodity can be interpreted as the market price of the commodity. The shadow prices of the balance constraints are reported by TIMES in the Par_CombalEM and Par_CombalGM parameters, shown under VEDA as the EQ_CombalM parameter.

See also:  The Dual Solution of a TIMES Model, Section  2.2.7 Commodity balance EQ(l)_COMBAL

[xiao.li8@mcgill.ca]

See the documentation, Part I, section 3.2.4 Marginal value pricing :

"We have seen in the preceding subsections that the TIMES equilibrium occurs at the intersection of the inverse supply and inverse demand curves. It follows that the equilibrium prices are equal to the marginal system values of the various commodities. From a different angle, the duality theory of Linear Programming (chapter 14) indicates that for each constraint of the TIMES linear program there is a dual variable. This dual variable (when an optimal solution is reached) is also called the constraint’s shadow price, and is equal to the marginal change of the objective function per unit increase of the constraint’s right-hand-side. For instance, the shadow price of the balance constraint of a commodity (whether it be an energy form, material, a service demand, or an emission) represents the competitive market price of the commodity."

Based on that, I would say that the shadow price of the balance constraint of a commodity can be interpreted as the market price of the commodity. The shadow prices of the balance constraints are reported by TIMES in the Par_CombalEM and Par_CombalGM parameters, shown under VEDA as the EQ_CombalM parameter.

See also:  The Dual Solution of a TIMES Model, Section  2.2.7 Commodity balance EQ(l)_COMBAL

Hi Antti,

I am confused again. Please see below for the shadow price for ELC:

Attribute Process 2019 2020 2030

Cost_Flo FT_TRAELC 10286.21 12841.46 19819.83
EQ_CombalM - 109333.2 466.77 477.25
VAR_FOut ElcAgg_Solar 4786.41 4786.41 4786.41
ElcAgg_Wind 21071.81 14204.58 18595.94
EN_GasCCAdv 4036.8 1984.81
EN_GasCTAdv 249.22
EN_Hydro_CHN-1 587.94 910.51 910.51
EN_Hydro_CHN-2 149.77 292.82 292.82
EN_Hydro_CHN-3 933.035 934.015 934.015
Util_Rep_Batt-LFP 148.25
Util_Rep_Batt-NMC 811 1408.93

Take 2020 as an example: The wind, solar and hydro are all zero cost, so my electricity price is decided by marginal units--GAS, and as shown below, the price of two gas generators are both below 100... then why the EQ_CombalM for ELC are ~500??

ExRes: P=EN_GasCCAdv
Attribute Commodity Region - 2020 2030

Cap_New - China 252411.39
Cost_Fom - China 0 0
Cost_Inv - China 19639.35 19639.35
Cost_NPV - China 268886.21
Cost_Salv - China 55.2
Val_Flo ELC China -86174.14 -22029.27
ELCCO2Ng China 0 0
ELCNGA China 44804.075 22029.27
VAR_Act - China 4036.8 1984.81
VAR_Cap - China 320.42 320.42
VAR_CapM - China 65.44
VAR_Eout ELCNGA China 4036.8 1984.81
VAR_FIn ELCNGA China 7715.26 3793.44
VAR_FOut ELC China 4036.8 1984.81
ELCCO2Ng China 431977.57 212394.82
VAR_Ncap - China 320.42
VAR_NcapM - China 456.29
VAR_NcapR - China 17.31
VAR_POut ELC China 193790.56 95270.65



ExRes: P=EN_GasCTAdv
Commodity Region Attribute - 2020 2030

- China Cap_New 113004.19
Cost_Fom 0 0
Cost_Inv 8790.83 8790.83
Cost_NPV 116721.81
VAR_Act 249.22
VAR_Cap 164.96 164.96
VAR_CapM 65.44
VAR_Ncap 164.96
VAR_NcapM 427.3
VAR_NcapR 101.26
ELC China Val_Flo -36264.84
VAR_FOut 249.22
VAR_POut 11930.083
ELCCO2Ng China Val_Flo 0
VAR_FOut 42362.91
ELCNGA China Val_Flo 4393.81
VAR_Eout 249.22
VAR_FIn 756.61


Did I miss something like the marginal demand ?

Best,
Xiao

[Antti-L]

Why addressing me? The shadow prices (marginals) are determined and reported by the solver, and so if you have doubts about what the solver reports as the dual values, you could approach the solver developers. 

But anyway, how many timeslices do you have?  If more than a single one (ANNUAL), I think you should have more clearly shown those results parameters for the individual timeslices. Nonetheless, from what you have shown, one can at least see the average weighted price of the produced ELC, by dividing the Val_Flo for ELC by the corresponding Var_Fout. For the first gas process shown I get 86174.14 / 4036.8 = 21.35 as that average weighted price in 2020. But the price of ELC at individual timeslices remains undisclosed.

[xiao.li8@mcgill.ca]

Why addressing me? The shadow prices (marginals) are determined and reported by the solver, and so if you have doubts about what the solver reports as the dual values, you could approach the solver developers. 

But anyway, how many timeslices do you have?  If more than a single one (ANNUAL), I think you should have more clearly shown those results parameters for the individual timeslices. Nonetheless, from what you have shown, one can at least see the average weighted price of the produced ELC, by dividing the Val_Flo for ELC by the corresponding Var_Fout. For the first gas process shown I get 86174.14 / 4036.8 = 21.35 as that average weighted price in 2020. But the price of ELC at individual timeslices remains undisclosed.

Hi Antti,

Thank you, these are helpful.

Best,
Xiao