GCAM7.0 errors: Equilibrium cannot be reached for a given 1.5 degree emission path scenario without fixing GDP #298
Replies: 4 comments 3 replies
-
|
Period 10: 2045 Period 11: 2050 |
Beta Was this translation helpful? Give feedback.
-
|
Note:
These are indicative of solver trouble. But they are generated as a side effect of "bad prices". Often, the solver is able to recover. It seems like in your case it did for the most part.
This one is curious, particularly because it shows up alone and in many periods. And likely is related to something off about the new Macro model. I'll dig in and see if I can figure out why. Note, that GCAM 7 does have a bit more trouble solving. Particularly deep mitigation scenarios. When running in target finder mode it is going to try many different "dispatch". I have let it keep trying unless it is clear it will not converge and not panic on first failure. It isn't necessarily the case the greater the carbon price the more likely it will fail to solve. It is more nuanced than that. |
Beta Was this translation helpful? Give feedback.
-
|
I read the introduction of the current GCAM solver on github, which talks about how the current solver is prone to be difficult to solve because the derivative of the Jacobian matrix is zero. Is there any plan to adopt a new solver to solve this problem? |
Beta Was this translation helpful? Give feedback.
-
|
How are you designing scenarios? What variables are you changing to design 1.5 degree emission path> |
Beta Was this translation helpful? Give feedback.
-
|
To test the model, I only added carbon emission constraints to the model for each period, where the emission trajectory is referenced to the P2 scenario (sustainability scenario) from the IPCC 1.5°C Special Report |
Beta Was this translation helpful? Give feedback.
-
|
I would guess that 1.5 is a very stringent target to hit without making additional scenario changes. One approach to make that target potentially easier to reach is to adjust the share weights on advanced technologies. For example, EV and FCEV techs in many transportation sectors have share weight trajectories that go to 1 by 2050 or even 2100. But in a 1.5 trajectory world, I would expect that there would be investments in charging/fueling infrastructure and in longer range, faster charging batteries. These would decrease barriers such as range anxiety, etc. Reducing these barriers could be reflected by setting those share weight trajectories to transition to 1 sooner. |
Beta Was this translation helpful? Give feedback.
-
|
Maybe that's the problem, I will try to use different scenario changes later |
Beta Was this translation helpful? Give feedback.
-
I want to explore the impact of emission reduction on global GDP under 1.5 degree path, so I give a global emission path and set the parameter of FixedGDP-Path to 0. However, the following error prompt occurs when solving the model:
Period 5: 2020
Error adding to demand in marketplace for: FoodDemand_Staples, region: South Korea, value: inf
Error adding to demand in marketplace for: FoodDemand_Staples-budget-fraction-trial-supply, region: South Korea, value: inf
Error adding to demand in marketplace for: FoodDemand_Staples, region: South Korea, value: inf
Error adding to demand in marketplace for: FoodDemand_Staples-budget-fraction-trial-supply, region: South Korea, value: inf
Error adding to demand in marketplace for: FoodDemand_Staples, region: South Korea, value: inf
Error adding to demand in marketplace for: FoodDemand_Staples-budget-fraction-trial-supply, region: South Korea, value: inf
Model solved normally. Iterations period 5: 3982. Total iterations: 5011
Period 6: 2025
Model solved normally. Iterations period 6: 840. Total iterations: 5851
Period 7: 2030
ERROR:Model did not solve period 7 within set iteration 10032
ERROR:Currently Unsolved Markets:
ERROR:Unsolved Part 1: Solvable Markets
ERROR:X, XL, XR, ED, EDL, EDR, RED, brk, Supply, Demand, Mrk Type, Market,
ERROR:42898.7 , 38984.3 , 38984.3 , -3893.32 , 8.00842 , 8.00842 , 0.099815 , 0 , 42898.7 , 39005.4 , Trial-Value, South America_NorthernGDP_Trial ,
ERROR:Unsolved Part 2: Unsolvable Markets Not Cleared
ERROR:X, XL, XR, ED, EDL, EDR, RED, brk, Supply, Demand, Mrk Type, Market,
ERROR:
Period 8: 2035
ERROR:Model did not solve period 8 within set iteration 10134
ERROR:Currently Unsolved Markets:
ERROR:Unsolved Part 1: Solvable Markets
ERROR:X, XL, XR, ED, EDL, EDR, RED, brk, Supply, Demand, Mrk Type, Market,
ERROR:51817.4 , 49444.5 , 49444.5 , -2356.57 , 16.2988 , 16.2988 , 0.0476452 , 0 , 51817.4 , 49460.9 , Trial-Value, South America_NorthernGDP_Trial ,
ERROR:Unsolved Part 2: Unsolvable Markets Not Cleared
ERROR:X, XL, XR, ED, EDL, EDR, RED, brk, Supply, Demand, Mrk Type, Market,
ERROR:
Period 9: 2040
ERROR:Model did not solve period 9 within set iteration 10101
ERROR:Currently Unsolved Markets:
ERROR:Unsolved Part 1: Solvable Markets
ERROR:X, XL, XR, ED, EDL, EDR, RED, brk, Supply, Demand, Mrk Type, Market,
ERROR:57087.5 , 52076.3 , 52076.3 , -5033.04 , -15.3256 , -15.3256 , 0.096688 , 0 , 57087.5 , 52054.4 , Trial-Value, South America_NorthernGDP_Trial ,
ERROR:Unsolved Part 2: Unsolvable Markets Not Cleared
ERROR:X, XL, XR, ED, EDL, EDR, RED, brk, Supply, Demand, Mrk Type, Market,
ERROR:
Period 10: 2045
ERROR:Model did not solve period 10 within set iteration 10087
ERROR:Currently Unsolved Markets:
ERROR:Unsolved Part 1: Solvable Markets
There are multiple markets that will not reach equilibrium by 2045. How should this problem be solved? Can this problem be solved by changing the parameters of the solver?
Beta Was this translation helpful? Give feedback.
All reactions