HandbookReference

Glossary

The LCA terms this handbook uses, and the words PrismLCA has for its own things — in alphabetical order, each with a link to the page that covers it.

Terms from A to Z#

Activity: Brightway's word for a process. PrismLCA says process; you meet the word activity in Brightway, in the Activity Browser and in ecoinvent's activity types. See Process.
Allocation: How the burdens of a process that makes more than one product are split between those products — by mass, by economic value, or not at all if you expand the system instead. You record your allocation rules in the model's Model info under Rules & allocation. Background datasets come with the allocation of their system model. See Cut-off system model.
Approve report: Your sign-off on a report: you have read the chapters the models wrote or changed, and the report says what you want it to say. A chapter that changes afterwards is marked New until you approve again, and the export asks about it. See Review and approve a report.
Audit log: The list of every change to a model — yours, the assistant's and an external agent's — with who made it. It sits on the Model, Parameters and Scenarios pages; changes on the undo stack can be taken back from there. See Check, undo and the audit log.
Background database: The large, read-only database of datasets you did not model yourself: electricity, materials, transport, waste treatment. PrismLCA works with ecoinvent (bring your own licence) and the free BAFU database. A workspace can hold one of them or both side by side. See Background databases.
Characterisation: The LCIA step that converts each elementary flow into its contribution to an indicator, by multiplying it with a characterisation factor — for climate change, kilograms of methane times its CO₂-equivalent factor. You can compare and edit factors on the LCIA page. See Impact methods.
Cut-off system model: One of ecoinvent's ways of modelling recycling and by-products: a process that recycles material carries the burden of the recycling, and the recycled material enters the next life cycle burden-free. PrismLCA's ecoinvent import expects the cut-off release, for example ecoinvent 3.12 cut-off.
Elementary flow: A flow between your product system and the environment — carbon dioxide to air, water from a river, land occupied. Elementary flows are what impact methods characterise. Results → Decomposition tree lists them. See Find the hotspots.
Exchange: One input or output of a process, with an amount: an arrow on the canvas. Exchanges link processes to each other and to elementary flows. See Build on the canvas.
Expert review: See Approve report.
Foreground: The part of the product system you model yourself, with your own data — the processes on your canvas. In PrismLCA each model is its own foreground database. See Key ideas in five minutes.
Functional unit (FU): The quantified function your study is about, which all results refer to — “one garden hose reel set used for ten years”, not just “a hose reel”. In PrismLCA you select one process as the functional unit; the toolbar shows it as FU: with its amount and unit, and Calculate needs exactly one selected. See Calculate.
Health check: A structural check of a model: missing units, unlinked references, circular links, duplicate flows, missing reference products, mass balance and unit dimensions. The assistant runs it when you ask for it. See Check, undo and the audit log.
Hotspot: A process, material or elementary flow that contributes a large share of an impact — the place where a change would matter most. The contribution tree under Results → Decomposition tree shows them. See Find the hotspots.
Impact method: A set of indicators with their characterisation factors, such as EF 3.1. In PrismLCA's menus an impact method is usually picked as a method family plus the indicators you want. See Impact methods.
Indicator: One impact category with its unit — climate change in kg CO₂-eq, acidification in mol H⁺-eq. You choose indicators on the Results page. See Compare results.
LCIA: Life cycle impact assessment: the step that turns the inventory — all the elementary flows of your product system — into indicator results, through characterisation. Also the name of the PrismLCA page where you compare and manage methods. See Impact methods.
Localize: Making an editable copy of a background dataset inside your model. Background datasets are read-only; right-click one on the canvas and choose Localize (edit) to change its inputs, for example to use your own electricity mix. The assistant can do the same, and can take the dataset's direct suppliers along. The original database stays untouched. See Find the right dataset.
Market dataset: A background dataset that represents a product as it is bought in a region: a mix of producers plus the transport and losses to get it to the customer. A production dataset is one producer's process, without that delivery. A market group combines several markets into a larger region. The search panel filters them with P, M and G. As a rule, link markets unless you know the producer. See Find the right dataset.
MCP: Model Context Protocol: the open standard through which AI clients such as Claude Code, Codex or Claude Desktop can use PrismLCA's tools. Off by default. See What MCP is for.
Method family: A group of related indicators published together, such as EF 3.1. PrismLCA's shipped families are read-only; you can build your own family from copies on the LCIA page. See Impact methods.
Model: One product system in PrismLCA: its own foreground database with a canvas, parameters, scenarios, goal and scope. A workspace can hold many models. See Build on the canvas.
My processes: Your library: one per workspace, holding datasets you imported from other software or moved there from a model. Models link to it live, so a correction in the library reaches every model that uses the dataset. See Your library: My processes.
Parameter: A named value — a mass, a share, a distance — that amounts and formulas can use. Change it once and every amount that uses it follows. See Parameters and formulas.
Plan card: The list of steps the assistant writes for a longer task and ticks off as it works, shown beside the chat. See Work with the assistant.
.prism bundle: A file with the extension .prism that holds one model and references to the background data it uses — not the background data itself. Use it to send a model to a colleague or move it to another computer. See Share and export models.
Process: A box on the canvas: something that turns inputs into outputs — producing a part, transporting it, using or disposing of it. Every process has one reference output. See Build on the canvas.
Reference: A background dataset placed on your canvas. It is a locked pointer to the dataset in the background database, not a copy, so you cannot edit it there — Localize it if you need to. One dataset can stand on the canvas several times, one box per tag its inputs carry. See Build on the canvas.
Reference flow: The amount of product needed to fulfil the functional unit — for one hose reel set, the reel, the hose and the fittings in their amounts. In PrismLCA it is the reference output of the process you choose as functional unit. See Calculate.
Release: The earlier name for approving: see Approve report.
Scenario: A named set of parameter values that replace the baseline values for one calculation — “recycled reel”, “rail instead of truck”. Scenarios appear on the Results page next to the model. See Scenarios.
Sensitivity analysis: How far the result moves when an assumption changes. The Sensitivity analysis tab under Results shows it as a tornado over the model's parameters and as a curve of one parameter across a range. See Test how sensitive the result is.
Skill: A Markdown guide the assistant can load for a kind of task, such as choosing a proxy or writing the goal and scope. PrismLCA ships ten; you can add your own. See Skills and tools.
Subject: Something you calculate and compare on the Results page: a model, a scenario, another model or a single dataset. See Compare results.
Tag group: A set of labels you attach to processes, such as life-cycle phase (production, distribution, use, end of life) with one tag per phase. The inputs your model buys in can carry a tag of their own too, such as raw materials for the steel of a manufacturing step. Results can be grouped by a tag group. See Life-cycle phases and tags.
Workspace: Your background data — ecoinvent, BAFU or both — together with the models built on it and your library. You switch, back up and delete workspaces at the bottom of the sidebar. See Workspaces, backups, starting over.