ecoinvent Forum Archive

Good morning,

I am modelling farms and have two questions regarding the fertilizer use. I'm using the 3.7 version.

1. Let's imagine that I want to model monoammonium phosphate (8.5% N, 52% P2O5) with an application rate of 100 kg/ha. The process "market for monoammonium phosphate | monoammonium phosphate | APOS, S, RNA" with Traci 2.1 results in 859 g CO2/kg. Thus, applying 8.5 kg N and 52 kg of P2O5 (100 kg/ha) would emit 85.9 kg CO2. However, the way nutrient supply is modeled in the databse asks you to separate N from P2O5. So if I want to apply 8.5 kg N, I would have to do 100/8.5=11.76 kg MAP/kg N and then multiply by 8.5 kg N, resulting in 100 kg of MAP. In GHG, 11.76*859=10105 gCO2/kg N and then *8.5=85.9 kg CO2 (same as before). If I then add the P2O5, which would be 100/52=1.92 kg MAP/kg P2O5 and then *52=100 kg of MAP, results in an additional 85.9 kg CO2.  Thus, total emissions calculated this way would be 171.8 kg CO2, double the initial result. The same applies to the rest of fertilizers that have 2 or more nutrients. Are we double counting? Where is the mistake in my reasoning? I would appreciate any help.
2. My second question is regarding the increase in emissions from N from monoammonium phosphate from version 3.6 to 3.7. In version 3.6, monoammonium phosphate, as N, results in around 3 kg CO2/kg N. In version 3.7, it is 10.1 kg CO2/kg N. Why is there such a difference?

Thanks!

Dear jaantona,

Thank you for your thread. The mistake in your reasoning (in my opinion) is that you take two allocated products, sum their score and then compare this total score with the product that is actually allocated to produce those two distinct products. If you know the amount of MAP used in your model I propose you to use the market for MAP and directly model this product. The nutrient supplies are there for (among other reasons) to help users model nutrients when they do not know where those nutrients are sourced. They do however know that they need an X amount of Nitrogen. In those cases, this is useful for the users.

Regarding your second question, the difference is due to many reasons. The datasets were updated and scaled. I do not recommend comparing the nutrient supply in v.3.7 with the dataset from v.3.6. It is not fair, it contains an extra and more accurate transportation burden. Not only the datasets were updated but also inputs to the dataset itself now refer to updated cases. Please have a look in our report of changes, the documentation helps a lot understanding the full picture of the update.

Please let me know if you have further questions.

Kind regards,

Avraam, Data analyst, ecoinvent Association