I dont't know because my research is focused on another topic

Thanks Antonio Alberto Clemente . I am not sure whether you see this question as you do not follow me. But I assure you I did not send any email regarding this question. I appreciate your response anyway :) Cheers.

even if the issue does not have specific publications on it I believe it is quite well covered in LCI databases that are integrated with LCA software. If you add a computer based workload to your investigated case you can track the environmental burdens that are related to it. there are also water specific measures (water footprint) that give only water related impacts as a results. see one of the milestone papers on it: https://s3.amazonaws.com/academia.edu.documents/45515724/Complementarities_of_Water-Focused_Life_20160510-31712-17dur6g.pdf?AWSAccessKeyId=AKIAIWOWYYGZ2Y53UL3A&Expires=1545152501&Signature=Q6RzFM%2BT74z4odGRMFAby0VF3jY%3D&response-content-disposition=inline%3B%20filename%3DComplementarities_of_Water-Focused_Life.pdf

Truly interesting question. An issue I've never thought about. Indeed, we spend resources when making our simulations. These are a cost for the environment. However, if you run a firm or your university spends electricity to feed your computer, this is already accounted for.

Find some reliable data about your computer electrical consumption (e.g. Wh/operation or kW at a fixed CPU usage). Then consider the electicity mix of the country in which your machine are operating and you roughly have it.

Maybe you should consider to split production and end of life of your machine into all the hours of work of your machines in its life.

Thank you Tomasz Nitkiewicz , Davide Tosches and Gianvito Scaringi

Water footprint of computer:

There are basically two approaches, Water footprint Network (WFN) approach and Life cycle assessment (LCA) community approach.

Selection of the method is based on your objectives. However, since this is more related to a product, I suggest LCA approach (which is based on ISO standards).

Behnam Ababaei , if possible please let me know your scope and objectives.

Anyway, in case you want to follow WFN approach, you need to consider basically the Blue water (water consumed for the production) and Gray Water (water required to assimilate the pollutants generated due to the production).

Amila

Dear Amila Abeynayaka . Thanks for the response. Actually what I need to do is to estimate the footprint of running simulations for one unit of time, not over the whole life cycle of a computer. I need to know how much footpring is left when these simulations are run. To simplify, I can assume that the CPU is 100% engaged during a unit of time (1 minute or hour) and how much energy/water/money/etc it consumes. Cheers.

Behnam Ababaei , sorry for not reading the question properly,

"...make an hour / a unit of mathematical calculations with a computer possible and how much emission is created..."

it seems your objective is to quantify the pollution load (such as carbon footprint, water footprints) for a unit run of your computer.

In that context, I agree with what Tomasz Nitkiewicz and Davide Tosches suggested. Energy consumption and computer life loss are the important factors to be considered (consumptions). These data available at standard LCA data bases like ecoinvent. then you may need a software like SimaPro (or openLCA) or simple excel program.

However, I would like to open up your objectives and scope bit more.

For example, lets see the possibilities of social and other gains. Some computers consume more energy but need more human assistance to complete the mathematical calculation. then lets say the operator has to switch on the screen and press a key. Some may not required. in such cases, there is social/health and other aspects to be considers (I understand your objective is not a research but a simple calculation), but then Why are you doing this? with that question, then we can agree with what Gianvito Scaringi said."firm or your university spends electricity to feed your computer, this is already accounted for".