Pickler has choosen for Idemat as their default LCI database across the whole application. It's important to only choose 1 LCI database, in order to enable benchmarking and prevent cherry picking for users.
The most important factors for Pickler decision is that calculations are created transparent, are up-to-date and science based. Since this is also required by upcoming legislation.
In the text below the main differences between the two LCI's are explained.
Idemat results compared to Ecoinvent (EI) results
Stets of LCA data are built for specific type of purposes (type of users), like indicator systems are (see FAQ 3.2 for indicator systems). Although the differences in indicator systems are bigger than in LCA datasets, LCA datasets have different characteristics for different user groups. This means in LCA benchmarking that mixing of datasets in one calculation must be done with great care. As an example: for plastics, the EI data are 10% – 30% higher than Idemat (on carbon footprint and eco-costs)
There a 6 reasons that calculation results in EI are higher than in Idemat (in order of importance):
a. Much of the LCI data in EI are based on old LCIs (or extrapolations of very old datasets), in contrast to Idemat scans the internet on recent publications
b. For electricity, EI is lagging behind
c. For transport the EI data are based on statistical data, in contrast to Idemat that applies measured data of modern trucks and vessels
d. EI includes ‘infrastructure’ (e.g. the steel in the chemical production plants)
e. EI does not have a cut-off point, where Idemat ignores subsystems with less than 1 or 2%
f. EI tends to apply the ‘cautionary principle’ by applying worst case scenarios
It is obvious that, because of the many measures to decrease carbon emissions in the recent past, point a. and b. lead to a quite serious overestimation: it can almost explain all differences for plastics. EI as well as Idemat use PlasticsEurope as source for their LCIs. Since PlasticEurope is frequently updating their data, the slow renewal by EI is causing quite significant differences.
With regard to outdated data on electricity see this website and the paper [Olindo et al 2021]
The issue in point c is that EI is based on quite old statistical data in the EU (HABEFA v3.1) on groups of transport vehicles , in combination with the maximum emissions of EURO 6, whereas Idemat applies measured data from road tests of standard modern trucks in the EU.
Idemat is based on the common practice in long haul transport: 50% full load (100% to client, empty back). The life span in Idemat is 1000.000 km (common in the transport business), instead of 540.000 km (!) in EI. In combination with the lower fuel consumption in Idemat, the result is that EI has more than 30% higher carbon footprint and more than 45% higher eco-costs.
Idemat is based on actual data of the ‘Emma Maersk’ containership. EI seems to base the data on a ship that is 5 times smaller, and fuel consumption before the ‘slow steaming’ practices, and new ship designs to reduce fuel consumption. The difference is a factor 2 in carbon footprint, and even a factor 5 in eco-costs, because of old sulfur dioxide emissions in EI.
An important difference between EI and Idemat, point d, is that EI includes ‘infrastructure’ (e.g. the steel in chemical production plants. This accounts for 2% – 4% extra ecoburden.
Point e results in a difference of approximately 2 %
Point f results mainly in applying ‘safe side’ assumptions. An example is that EI applies quite often excessive SO2 emissions, assuming the maximum allowable sulfur in oil based fuels, but ignoring the additional requirements in local environmental nuisance requirements. Another example is the assumption that all potential pollutants in coal that are burned in coal fired power plants will enter the environment (via the chimney), whereas all modern coal fired power plants have exhaust filters.
Another issue is the fact that EI does not count the combustion with heat recovery (ISO 14044, section4.3.3.1.), since they claim that this would cause double counting. This is right for country wide calculations, but not true for marginal calculations on product innovation. See FAQ 2.4 Note 1.
Conclusions
1. when you make ex-post LCA calculations, you might prefer the Ecoinvent dataset, since it follows the ‘precautional principle’ and you are ‘safe side’.
2. when you make ex-ante calculations on innovations of products and services, you run the risk with de Ecoinvent dataset that you take the wrong decisions since you apply outdated data.
Note 1
Some people like to exaggerate LCA related issues. However, the policy of overestimation of LCA results is a bit dangerous, since that may lead to the wrong decisions, and often does not help much in convincing people.
Note 2.
More on the specific calculation rules of Idemat can be found at this website.
Differences in transport footprint between IDEMAT and EcoInvent
In most of the standard software packages, the data on transport are only given in the unit “ton.km”. The reason is that all standard LCI databases (like Ecoinvent) only supply data on the basis of tonnes x km. It is, however, good to realise that the LCIs are calculated on the basis of a full load of the truck (or vessel, or plane) and an empty trip back, divided by the maximum load of the transport vehicle. When the density of the freight is relatively low, the truck is full at a maximum volume instead of a maximum weight. In such a case, a correction factor has to be applied, since the energy required for long distance transport is dependant on distance, shape and velocity, and hardly dependant on the weight.
The correction factor must be applied when the density is lower than:
– 160 kg/m3 for airfreight
– 320 kg/m3 for freight in a European standard truck + trailer
– 414 kg/m3 for freight in a standard truck + container (40 ft)
– 843 kg/m3 for freight in a standard 20 ft sea container (take this density for other sea freight as well)
The correction factor to be applied is “break-even density” / “actual density”
under the condition that this factor is more than 1.
Then, the amount of ton.km for the input of Simapro or CES has to be calculated as follows:
“actual tonnes” x “actual km” x “break-even density” / “actual density”
Example: when 24 tons has to be transported by a standard European truck and trailer (24 tons = a full truck load for high densities), and the actual density is 160 kg/m3, the correction factor is 2. This means that the truck must drive two times to transport this freight. The eco-burden per ton.km of this transport is 2 times the eco-burden per tkm of high density freight.
See also Section 4.1 of the Practical LCA Guide
Note 1:
The assumption that the average load factor (=occupation) is 50% (the truck is full, but is empty back, on average) is realistic in practice. It is obvious that, from environmental point of view, it must be avoided that the truck is not fully loaded. If this is not the case in a typical situation, a multiplier must be applied in LCA to cope with the partly loaded truck. When, in special cases, the trip of the truck can be combined with other freight on the trip back, the so called “economic allocation” of the eco-burden of round trip of the truck must be applied (which will result in an multiplier less than one for the ton.km).
Note 2:
There is a fundamental difference between the Idemat calculation of transport by truck, and the calculation of Ecoinvent:
– Ecoinvent calculations are based on (1) the metric tons gross vehicle weight (GVW) (2) the average load factor per country (3) the EU norms on fuels and emissions
– Idemat calculations are based on (1) the maximum net load of a truck (2) the assumption “100% full on the way to the client, empty on the way back” (3) the fuel consumption as measured in real road tests of commercial Euro 6 trucks.
Therefore, Ecoinvent is more suitable for policy calculations in countries, whereas Idemat is more suitable for calculations on production chains.
Note 3:
The Idemat data on sea freight and air freight are based on typical aircraft and sea vessel data and actual load factors. The Ecoinvent data are based on statistical averages. The Ecoinvent data on transport by sea and by air seem to be a bit outdated.