Forum for ecoinvent Version 3

Here you find all boards with the latest posts of the Forum for ecoinvent v3

Written on 23.02.2018 by luciavalsasina see context

Dear Johanne Hammervold, 


the inventory of tovex has not be significantly changed since version 2 and it still refers to the documentaiton in verison 2 report number 7 on Building Products. 


If you are seeing something different, pelase let me know so that I can investigate it. If you think a screenshot might be helpful, feel free to email me at


Best regards,


Lucia Valsasina, Data Analyst, ecoinvent

Written on 23.02.2018 by Florentine Brunner see context

Dear Miguel and Lukas,


The standard heating value for natural gas of 39 MJ/Nm3 refers to the Higher/Gross Heating Value. Although this is meant to be the default assumption for all datasets producing or using natural gas, you are right that there is a number of datasets in which the property “heating value, net” is set to 39 MJ/Nm3. I agree this is confusing and we are therefore doing our best to correct the issue and provide more explicit documentation on the matter.


Values for the density of natural gas originate from ecoinvent version 2, where a density of 0.84kg/Nm3 was defined for sweet gas and of 1 kg/Nm3 for sour gas (see Tab. 3.4, p. 6 in Faist Emmenegger et al. 2007, pdf report 06_V_Erdgas accessible in the report section of ecoinvent version 2). With the standardization of the gas quality to (sweet) gas of 39 MJ/Nm3 gross heating value, the density was standardized to 0.84 kg/m3. The density of 1kg/m3 in the dataset “natural gas production, DE” was hence mistakenly retained from before the introduction of standardized gas quality and stems from the sour nature of natural gas in Germany.


Reference: Faist Emmenegger M., Heck T. and Jungbluth N. (2007) Erdgas. In: Sachbilanzen von Energiesystemen: Grundlagen für den ökologischen Vergleich von Energiesystemen und den Einbezug von Energiesystemen in Ökobilanzen für die Schweiz (ed. Dones R.). Swiss Centre for Life Cycle Inventories, Dübendorf, CHKind regards,Florentine BrunnerData Analyst, ecoinvent

Written on 23.02.2018 by Anders Damgaard see context

I was wondering if anyone have taken a look at Kraft Paper production, RER. I am a bit surprised about the results when comparing bleached versus unbleached production. In most categories bleached comes out worse than unbleached, but when comparing on basis of ILCD categories the following categories come out better for bleached: Human-Tox, non-carcinogenic; Depletion of abiotic resources (elements); Ecotoxicity; and Freshtwater Eutrophication.

To me this is rather counter intuitive, as the initial paper production steps should be the same besides the bleaching of the final product. A part of the reason is the data quality, where the dataset for bleached paper is not as “rich” in included data as unbleached paper. But I can’t figure out if this is also something with the linked processes between them.


I also tried to compare in SimaPro (the others were done based on downloads from Ecoinvent directly) and here the results are even more weird as I get savings in a number of categories when considering consequential modelling, which are directly opposite to the results from the Ecoinvent website. For now my proxy will be that bleached must be equal to or worse than unbleached in all categories. But if anyone can give hints to reasons I would be grateful.





Written on 23.02.2018 by jrsfinn see context

Dear All,


I'm not fully understanding how elementary flows operate within the datasets - may you please help?

I'm modelling a decorated glass build process. Within this process, I require Stannane (or Butyl trichloro tin) to represent the amber decorated coating applied to the clear glass bottle. 

The only available flow for this is elementary and I'm not sure how it operates within the process, as I can't choose a provider, etc. I know these flows are the base of all processes when you get deep down into the data, but I've never used one like this before. Will it contribute to the impacts as it's supposed to?


Please offer help! 

Written on 22.02.2018 by lukasl see context

Thanks Florentine for the answers and thanks Miguel for the hints. I would be also interested in the source for the density of 0,84 kg/m3 as well as the standard lower heating value and if they are consistent within the database. I was not able to find it in the ecoinvent dataset documentation.


Thank you very much.


Written on 21.02.2018 by Hammervold see context


The inventory for the process "Explosive, Tovex" has changed quite a bit in ecoinvent 3, and I can't find any documentation of this new process. 

Is this available anywhere?

Best regards

Johanne Hammervold

Written on 16.02.2018 by Miguel - F.Ast see context

Dear Florentine

where does the 0.84 kg/m3 value come from? I checked the UPR description of the process and I have not found a reference to the density.

You mention that the heating value of natural gas is standardised in ecoinvent to 39 MJ/m3. Does it mean all the datasets assume this value?

By chance I was checking the natural gas use "electricity production, natural gas, combined cycle power plant" in CA-ON and the natural_gas_net_heat_value is 39 MJ/m3. The comment specifies that is the lower heating value measured at 0C 1 atm. Other datasets refer to gross heating value of 39 MJ (without specifying the reference conditions of pressure and temperature).

is the 39 MJ (gross) heating value a standard and thus the reference to net heating value an error of the CA-ON dataset? or do different datasets have different heating values ? that would imply that there's no standard heating value for gas in ecoinvent.

thanks for the clarifications! Miguel

Written on 16.02.2018 by Florentine Brunner see context

Dear Miguel,


While for natural gas the standard heating value of 39MJ/m3, different from specifications in the data quality guidelines, was introduced this is not the case for other fuels such as petrol so far.



Florentine Brunner, Data Analyst, ecoinvent

Written on 16.02.2018 by Florentine Brunner see context

Dear Miguel,

thank you for pointing this out. The density of natural gas specified in the dataset ‘natural gas production, DE’ should indeed be in line with the density of 0.84 kg/m3, which matches the standard heating value of 39 MJ/kg.


Florentine Brunner, Data Analyst, ecoinvent

Written on 15.02.2018 by luciavalsasina see context

Belay Teffera,

The transport dataset in ecoinvent is expressed in ton*km. Therefore, if you have information regarding the weight of the goods transported and for how many kilometres it is transported, you can use it regardless of the fact that the pieces are transported together or separate. What you want to consider is the size of the lorry you pick depending on the weight and size of what you are transporting, you can read all necessary information regarding lorries in the general comment of the datasets. You can access the information logging in ecoinvent online, if you never logged in please read here.

Said that, what is modelled in ecoinvent are conventional lorries. I am not an expert in wind turbines, but I assume that, at least for certain parts, specific vehicles might be required. If transport is a key aspect in your study, then you should look into modelling transport according to the specific requirements of wind turbine transport. I would advice you to consult literature to evaluate the importance of transport. Please mind that ecoinvent is a background database.

Best regards,
Lucia Valsasina, Data Analyst, ecoinvent

Written on 14.02.2018 by Miguel - F.Ast see context

what is the reference higher heating value of petrol in ecoinvent datasets (if existing). I've seen other fuels have it defined in the properties (e.g. natural gas). For the case of natural gas, the reference value is 39 MJ (/m3?), different from the value of the data quality guidelines (40.3).

I have not seen the value specified for petrol, and since there's already a discrepancy for natural gas, I wonder if the value provided in the data quality guidelines is the best reference.


Written on 14.02.2018 by Miguel - F.Ast see context

another way to look it is emissions. I've been checking this and for other datasets and if you assume total oxidation of the fuel, you get a very good aproximation of CO2 emissions.

in the case of combusting natural gas, if density was 1kg/m3 it does not work.

take the Swish process "

heat production, natural gas, at boiler condensing modulating <100kW

" as an example.

it emits 2.18 kg CO2 per m3 of NG. NG is reported to have C content of 69% by mass. if 1 m3 was one kg, then we'd combust 0.69 kg of C per m3. For each 12kg C we have 44 Kg CO2, that is 2.53 kg CO2/m3 combusted. To be consistent with emissions density would need to be 0.86 kg/m3. Something does not work.

Written on 14.02.2018 by Miguel - F.Ast see context


Are you sure about the density of 1kg/m3? I may be wrong but crunching some numbers that does not make sense to me.

CH4 has a density of 0.67 kg/m3 at 15 degC and 1 atm (the conditions used as reference in ecoinvent data quality guidelines).

C3H8 (propane) has a density of 1.86 kg/m3 at the same conditions

if we assume NG is only methane and propane, to have 1 kg/m3 you would need to have 72% of methane by volume, which is very low. usual ranges are 87%-97%

I did some back on the envelope calculations and I think that to have 39 MJ/m3 it would need to be 97.6% methane by volume, which seems more reasonable.

could you please confirm the density value?

Written on 12.02.2018 by belaytef see context

Dear ecoinvent team

In wind turbine, nacelle is an assembly of subassemblies/parts such as baseplate, gearbox, generator, etc. In some case, for transporting, it may also include or be connected to the hub, which is actually part of the rotor of the wind turbine.

In my case, the supplier of the wind turbine components and the client (developer) are in different countries. Transporting of these assembles/subassemblies involve different transport modes such as road and sea transport

Is there any difference in the resulting transport datasets between transporting 'each subassemblies / parts' separately and 'the assembly', knowing approximate weights of each subassemblies/parts and approximate distances for each transport modes?

Written on 25.01.2018 by Florentine Brunner see context

Dear Lukas,

thank you for your question. You are right that the exact conditions for natural gas networks at different pressure levels are not specified in ecoinvent. These will differ from system to system and with the gas compostion. We therefore recommend to scale the amount of natural gas required for your system based on itsheating value. The heating value for natural gas in ecoinvent is standardized to 39MJ/m3. Moreover, what you can find in the properties of the mentioned dataset is the dry mass of the ´natural gas, high pressure´, which in this case is 1kg/m3.

I hope this answer serves you.
Kind Regards,

Florentine Brunner
Data Analyst at ecoinvent