SM5002 Chap.2 Collecting Life Cycle Inventory Data
Collecting Life Cycle Inventory Data
What the inventory phase produces
The life cycle inventory is the second phase of an LCA. ISO 14044:2006 describes it as the compilation and quantification of input and output data for the product system across its whole life cycle, collected to meet the goal and scope already defined. The result is a long list of flows drawn from and released to nature, each with a quantity attached.
Lecture 1 sets two outcomes for this phase: identifying suitable data sources and appraising data quality.
Three kinds of activity data
The GHG Protocol standard for product life cycle accounting distinguishes three types of data. Direct emissions data records what a process releases or removes, established by monitoring, stoichiometry, mass balance or similar methods.
Financial activity data is a monetary measure of a process that leads to emissions. Process activity data is a physical measure: the type and quantity of materials and products, the type, source and amount of energy, and the mode, fuel and distance of transport.
In a process-based LCA most rows are process activity data, which openLCA then links to an Ecoinvent process.
Primary and secondary, foreground and background
Two further questions classify every data point. The first is provenance. Primary data is collected first-hand for the specific study from measurements, internal records or surveys, and tends to be more relevant and accurate.
Secondary data comes from existing datasets or literature, such as Ecoinvent or scientific papers, and is easier to obtain but may lack specificity. The lecture contrasts the actual fuel use of a hospital sterilizer with a typical figure for sterilizers in general.
The second question is position in the system.
Foreground data describes processes under your direct control or influence, should reflect real operations, and comes from direct measurement, company records, supplier surveys and interviews, standard operating procedures, product specifications and technical data sheets, bills of materials, logistics records, simulations or reverse engineering.
Background data reflects industry averages, fills supply chain gaps upstream or downstream, and usually comes from inventory databases, where processes may be adapted for a different energy mix, transport mode or distance.
The two questions are independent: sterilizer electricity taken from a published paper is still foreground data, just secondary.
Pragmatic limits and the process map
Real inventories are shaped by constraints as much as by method: time, access to data, restrictions on reporting, proprietary information that may need a data sharing agreement, funding for primary collection, software and database licences, and the accuracy the study needs.
Financial data is routinely collected, as are energy, water and gas data, but often not at the resolution an LCA requires, so a mix of sources is normal.
Collection starts with a process map.
Engage stakeholders and verify with them, sketch a rough flow chart, learn the system and its terminology, trace the supply chain through its tiers, fix where each unit process begins and ends, identify intermediate, final and co-products, and list inputs, outputs, upstream and downstream processes.
A data collection sheet then asks, stage by stage, for materials, energy, water, emissions and waste, use phase details such as the number of uses of a reusable item, transport distances and modes, and end-of-life routes.
Effort goes first to likely hotspots such as the heaviest material or the most energy-intensive process, and estimates or input-output data are acceptable placeholders until a hotspot analysis shows where detail matters.
A data collection sheet in practice
The lecture closes with an exercise to design a collection sheet for a cradle-to-factory-gate LCA of single-use surgical scissors.
For raw material acquisition it asks for the metal type, the weight of one item and the address where the raw material is extracted. Transport to the manufacturer needs the vehicle for each leg, from lorries banded by size and emission class to transoceanic ships, air freight or freight trains.
At manufacture it asks for the site address, the number of instruments produced each year, annual electricity, water use and treatment type, chemicals, fuels such as natural gas, and the material and weight of individual and bulk packaging, including how many items each bulk pack holds.
From spreadsheet to software
Inventories usually begin in a spreadsheet and move into LCA software such as openLCA, SimaPro, GaBi or CMLCA, which in turn draw on inventory databases including Ecoinvent, the World Steel Association and ESU World Food.
Exercise 2 practises exactly this step with inventories taken from peer-reviewed papers on gloves, staplers, head covers and face masks, each row naming the Ecoinvent process to use.
Transport rows are entered in kg multiplied by km, calculated as the distance of each leg times the carried mass of product and packaging.
Quality, validation and gaps
Data quality is judged on five indicators: technological, geographical and temporal representativeness, completeness, and reliability.
Checks include consistency with other literature, whether inputs and outputs are proportionate, whether the scale makes sense, and careful verification of units to avoid conversion errors. Validation can use a mass balance, as with 205 g of PVC granules becoming a 200 g mask and 5 g of scrap, or a cross-check against secondary data, as with a decontamination machine measured at 10 kWh against 8 kWh in its specification.
Consistency, in the sense used by the ILCD Handbook, asks whether methods fit the goal and scope and are applied uniformly across the system, background datasets included.
When data is missing the lecture offers a sequence: identify the gap and its importance, use a proxy with the closest match, estimate or model it, ask stakeholders, use literature values, document every assumption, and run a sensitivity analysis to test robustness.
Inventory analysis is iterative, and each pass through these steps tightens the model.
What this chapter covers
- 01
Direct emissions, financial and process activity data
- 02
Primary against secondary data
- 03
Foreground against background data
- 04
Process maps and data collection sheets
- 05
Entering an inventory with Ecoinvent processes and transport in kg km
- 06
Data quality indicators, validation and filling data gaps
Worked example · free
Transport rows for a pair of nitrile gloves
- 1Find the carried mass in kg, including packaging.
- 3Multiply each leg's distance by the carried mass.
- 1Sum the legs and say why each keeps its own process.
Key terms
- Life Cycle Inventory
- The compiled and quantified inputs and outputs of a product system across its life cycle.
- Process Activity Data
- Physical measures of a process, such as material mass, energy use or transport distance.
- Primary Data
- Data collected first-hand for the specific study through measurement, records or surveys.
- Foreground Data
- Data on processes under the practitioner's direct control, intended to model real operations.
- Background Data
- Industry average data, usually from an inventory database, used to fill supply chain gaps.
- Mass Balance
- A check that total mass entering a process equals mass leaving plus any accumulation or loss.
Collecting Life Cycle Inventory Data FAQ
Is a figure taken from a journal article always background data?
No. Background and foreground describe where a process sits in the system, not where the number came from. Sterilizer electricity reported in a paper still describes a foreground process; it is simply secondary rather than primary.
Why does packaging mass go into the transport calculation?
Freight burdens scale with the weight actually moved. A box travels with the product on every leg, so leaving it out understates each lorry and ship row, even though the box later leaves by its own recycling route.
What should I do when a value in a published inventory looks wrong?
Run a sense-of-scale check against similar items, confirm the units, and go back to the paper or the data holder. Record the decision as an assumption and test the value in a sensitivity analysis.
Which data quality indicator does choosing a national electricity grid address?
Geographical representativeness. Linking manufacturing electricity to the grid of the producing country, such as China for the nitrile gloves, makes the data reflect where the activity physically happens rather than a global average.
Assessment move
Take one Exercise 2 inventory, label every row as primary or secondary and foreground or background, and recompute its transport rows by hand before entering them in openLCA. Then apply the five data quality indicators to the product you choose for the written assignment and list the gaps you expect to fill with proxies.
