How to solve challenges with Social-Life-Cycle Assessment - fitting database?

Question

We are currently exploring how to conduct social life cycle assessment with the help of openLCA on product scale. The goal is to identify design options for improving sustainability impacts (aka mitigate/ reduce them). 

Data we are using is country related, i.e. we derive risk values from that (e.g. for corruption). However, if associating such social impact flows with processes, these scale of course with the reference unit, which disvavours processes with large values in there respective reference units and selecting the unit already has a huge impact (e.g. changing from g to kg already changes the scale by factor 1000). This is not helpful for us and we struggle with the normalization/ weighting.

The questions are:

• which database is better suited for product scale, e.g. associating impacts with processes on a product scale to allow trade-offs aside from simply switching between countries (e.g. choising component A over B, because it involves less resources, which are associated with e.g. child labor risk); I tend to think, fromt he description, SOCA seems to be the better fit. But what about PSILCA?
• Are there databases, which provide information akin to "risk of X occurences of child labor per kg of product" for different processes? Or are all of them on a country level? 
• is it possible to generate one model of a product and do S-LCA and LCA (possibly LCC) with the same model? Or does that need adjustment/ new models for each purpose? 

Answer 1

Hi, interesting questions -

there is of course the soca database which is based on ecoinvent and thus reflecting products, but the main social LCA databases are based in IO databases that use money terms for all economic flows that link processes. This is not a limitation though. When you make a social LCA model, you should have a foreground system, based on products, and you can link this foreground system with a social LCA background database by setting a price for each product that is input in your foreground system. So e.g. if you have an input of 15 kWh electricity, you determine the price per product amount for this, and link this to an energy generation sector in the social LCA database.
This way you will not have a different result if you have an input product in 1g or 1kg as the price will reflect this change of amount.
This maybe answers your first two bullet points. Regarding the question of the combined model, this is possible with soca (if you are fine with the limitations - the extent for social LCA is somewhat reduced compared to e.g. PSILCA since the models are much smaller and do not contain services and only a small section of the overall economy).

Best wishes,

Andreas