Veda2.0 Released!


ObjZ Question
#16
Thank you for your response, Amit. I am looking at table 15, however, I do not see an option for reporting an annualized objective function value (ObjZ). I must admit this response may be a bit over my head. Is it possible to report an annualized present value of the objective function in VEDA BE? The reason I ask is that I would like to compare a business as usual scenario to a scenario with a carbon tax, and then calculate the marginal cost of abatement for CO2. The only way I can think to do this is to compare the non-tax costs of the two scenarios and the emissions between the two scenarios. To do this, however, I would need a non-discounted value for the objective function (without setting the discount rate = 0, which changes results drastically). Annualizing the objective function value I think would could help me solve this problem.

Thank you for your help, Mike.
Reply
#17
Apologies; there was some extra information in my last post that drowned the answer to the original question - "how to implement $SET ANNCOST LEV" in VFE.

short answer is to introduce the table shown below in SysSettings file

   
Reply
#18
hi Antti and Amit,

I am working with the EPAUS9rT model and have a question. I impose a $100 per ton CO2 tax starting in 2015 through 2050. The tax in 2014 is 0.

I look to VEDA_BE's "Cost_Comx" so show me the cost of the tax (revenue of tax) in each milestone year. The Cost_Comx should be equal to total CO2 emissions (Mtons) * $ per ton = Cost_Comx right? 

I ask because in 2015 cost_comx is 2,604,390 (M$) and CO2 emissions are 4913 Mtons. Shouldn't Cost-Comx be $100 * 4913 Mtons emitted = 491,300 M$? Something similar happens in 2020, but 2025-2050 the cost_comx is very close to $100 * emissions (see image).

Thank you,

Mike


Attached Files Thumbnail(s)
   
Reply
#19
You are not giving sufficient information.  
1) Earlier in this thread it has been several times mentioned that if you use the switch $SET ANNCOST LEV, then the annual costs are levelized, ortherwise they are the costs at the Milestone years.  Cost_Comx is one of the annual cost components, and so this difference applies to it as well.  It should be obvious that Cost_Comx can not be equal to emissions * tax rate in both cases. Which switches do you use?
2) You are not showing your data.  Please show the data on COM_TAXNET from VEDA-FE Browse.
3) You are not showing the emissions and Cost_Comx reported in VBE. Please show the VAR_Comnet and Cost_Comx values from the VEDA BE results.
Reply
#20
Thank you for your message Antti.

1. I am only talking about costs in the milestone years. Under a $100 tax, it is my understanding that if emissions at a milestone year are Y then the damages (cost_comx) in a milestone year should be $100 * Y. It seems that this may not be a correct assumption...If cost_comnet is an annual cost component then why shouldn't it line up with the cost in a milestone year?

I am not using any switches to levelize costs - I am just trying to understand the relationship between cost_comx and an emissions tax when there is a CO2 tax of $100 per ton.

2. Attached is output of emissions and Cost_comx and VAR_Comnet from VEDA_BE with CO2 in million tons and cost_comx is in million dollars. I have also attached a screenshot of the VEDA FE Com_Taxnet which shows $100 across the sectors.

Thank you for your help,

Mike


Attached Files Thumbnail(s)
           
Reply
#21
Ok, you say "I am only talking about costs in the milestone years."
But earlier in this thread you specifically wanted to know how to activate the switch $SET ANNCOST LEV, so I was assuming that you have now been using that setting. Is that correct or not?  And how about all the other switches related to cost accounting and the objective function? For example, do you still use $SET OBJ AUTO and $SET OBLONG YES?  When asking for help, provide all relevant information to help you.

Basically your assumption, that if the amount of emissions in a Milestone Year is Y and the tax rate is TX, the total annual taxes reported should be Y*TX, is wrong. It would only be correct in the case where you would not use the switch $SET ANNCOST LEV. Unfortunately, I still have no firm information about your switches.

I am also actually not seeing the Cost_Comx results from VEDA-BE in your screenshots, by region and commodity. You only give a picture of some Excel sheet.  And neither can I see the emission results by region and commodity (in the screenshot you show aggregate results for all regions and all commodities).

But the bottom line here is that I am not able to reproduce your issue. However, I would be glad to investigate your issue down to the bottom if you provide me with a reproducible case. If you cannot reproduce the issue yourself with any small test model (e.g. the DEMO model), I suggest that you provide me with the full input files of the model that you are working with right now. Just upload all the model input files and the listing file from the run case (*.DD, *.RUN and *.LST, zipped) to some server (e.g. Dropbox) and send me the download link in a private message. It would be easy to resolve your issue then, and would save a lot of time.
Reply
#22
In fact, I have been granted access to the EPAUS9rT model, v16.1 (an earlier development version).

I just tested with that version, defining COM_TAXNET for the sectoral CO2 emissions in all nine regions: Zero in 2014 and 100 USD/tonne in 2015.  And guess what:  When I divide the Cost_Comx results by the emissions results in each period, I am getting exactly 100 for all Milestone years 2015, 2020, ..., 2050. This is fully as expected and correct. (I was of course not using $SET ANNCOST LEV).

Thus, I am not able to confirm any of your findings even with the EPAUS9rT model, v16.1.  Consequently, I think the only way for me to see and explain what is going on in your model would be to have access to the full set of your model input files. What do you think?
Reply
#23
hi Antti,

Thank you so much for your help! I have sent you some screen shots as well as the .dd .lst .run files for the run EPAUS9rT_SCC_100. The .dd file has a _ts at the end of the file but I think it is the file you requested.

I have not turned on $SET ANNCOST LEV yet and I do not see the option for it when I open the control panel in VEDA FE. I have selected the OBJ Function Variant "Auto" and "Oblong" which I understand to mean that I have turned on switches for $SET OBJ AUTO and $SET OBLONG YES.

I think I may see what the problem is. Although the boxes for the SO2, NOX, and PM25 scenarios are not checked when I solve, they still show up when I select COM_TAXNET in the TIMES VIEW under browse. Does this mean that VEDA is still imposing the SO2, NOX, and PM25 taxes? If so, I am not sure why and how to turn them off.

Thanks again for all of your help. Please let me know if there are any other files you would like me to send. I am pretty sure the dropbox files I messaged you are the ones you requested.

Mike
Reply
#24
Thank you for the Dropbox link, but I asked for all the model input files (*.DD, *.RUN), so that I can run the model.
The following files were missing from the Dropbox folder:

base.dd
tradeparams.dd
syssettings.dd
baseextra-offeps.dd
growth.dd
refemi-ts.dd
uc-biofuelssimple.dd
uc-elc.dd
uc-buildings.dd
uc_trnhdv.dd
uc_trnldv.dd
uc-ind.dd
co2_tax.dd
uc-rpszeros.dd
Reply
#25
Antti,

I have added those files.

Mike
Reply
#26
I ran your model with the input files you provided, and the results look closely similar to yours.

However, I cannot see any problem with the Cost_Comx results with respect to the Var_Comnet values (emissions).  In the results, the ratios of Cost_Comx to the Var_Comnet values is always exactly 100 for all periods 2015, 2020, ..., 2050, for all regions, and for all of the emission commodities that have the COM_TAXNET=100 defined.  See a summary picture below:
   

As you can see, the total Cost_Comx value is about 2.6 million 'USD05' in 2015 and the sum of the Var_Comnet emissions are about 26,000. Thus, the ratio is exactly 100 as it should be according to the COM_TAXNET=100 defined.  It seems that you have just ignored USPCO2 in your results processing.  It may also be that your unit for UPSCO2 is actually different from the other CO2 commodities (kt vs. Mt), and if so, your tax rate would be wrong for UPSCO2.
Reply
#27
hi Antti,

Wow thanks so much for your help! You are right - either I accidentally deleted it or the USPCO2 was not included in the original EPA predefined CO2 Emissions table. The units are also in kt not Mt which means I will have to specify a different tax for UPSCO2.

Thank you so much,

Mike
Reply
#28
hi Antti,

I talked to the EPA and they confirmed that they accidentally left UPSCO2 in the model. So I will not be including it in any calculations in the future. I have another question that myself and EPA are somewhat stumped by. I am running two scenarios and in the BAU there are no REFCO2 emissions (refineries) but in the carbon tax there are REFCO2 emissions. Oil is refined in both scenarios so it seems there should be emissions in both. The EPA can confirm this is not a normal result but cannot help with why this may be happening. Do you have any suggestions?


Attached Files Thumbnail(s)
       
Reply
#29
Sorry, with such limited information, I have no other suggestions than to try and investigate the problem by yourself. In general, I think you cannot expect anyone else to do the debugging of your models for you.  

In other words, it sounds like you probably just have some bugs in your model. I would certainly be happy to help you, if my free "quality time" to be spent on helping others would be unlimited. Blush

However, if you have specific technical questions about TIMES, or can show some evidence of a possible problem in the TIMES model generator, I will certainly try to help.
Reply
#30
hi Antti,

Thank you for your reply. I have a quick VEDA BE question. I am looking at 2 scenarios: BAU and BAU with the levelized cost option turned on. I have turned on the levelized annual costs switch by adding the "$SET ANNCOST LEV" table to the SysSettings file. When I take the reported values from VEDA BE for each period from 2010-2050 and discount back to calculate the NPV for ObjZ I get exactly the same number: 51,279,034 reported by VEDA BE.

What I am wondering is why I get the exact number before subtracting the salvage value costs in my calculation? Does the levelized annual costs option somehow already include the salvage value? I have attached the results table from VEDA BE and the simple excel sheet that I used for this calculation. 

Thank you,

Mike


Attached Files Thumbnail(s)
   

.xlsx   Meeting 8 21 18 New.xlsx (Size: 49.05 KB / Downloads: 3)
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)