.png?width=200&height=58&name=Primary_White_XYCLE_Logo.png-09%20(2).png){kind=logo}
Uncertainty Analysis
What is Uncertainty Analysis?
Uncertainty analysis characterises the spread of possible outcomes in an LCA result, given the inherent variability and quality limitations of your input data. XYCLE uses your Data Quality Assessment (DQA) scores to derive the uncertainty of each inventory item, then runs a Monte Carlo simulation—sampling 1,000 times across all exchanges—to produce a distribution of results for each impact category.
The output is a histogram showing the range of plausible outcomes, alongside key summary statistics. This supports ISO 14044-compliant interpretation and is particularly important when making comparative assertions or submitting to critical review.
Why use Uncertainty Analysis in XYCLE?
Uncertainty analysis has traditionally required practitioners to leave their LCA platform and work in spreadsheets or code environments. Running it externally creates separate model instances that quickly fall out of sync with the primary model—any change to the LCA has to be replicated manually across each.
XYCLE runs uncertainty analysis directly from your existing DQA data, keeping everything in one traceable workflow.
Before You Begin
Uncertainty analysis requires a complete DQA—every eligible inventory item must have all criteria rated, including Completeness. The feature is locked in the Visualisation tab until this is done.
If you have not yet completed your DQA, see Applying Data Quality Ratings (DQR) for guidance.
Preparing Your DQA for Uncertainty Analysis
Two additions to your DQA feed directly into the simulation.
Completeness
A Completeness criterion has been added to both the Foreground and Background sections of the DQR panel. Rate it on the same 1–5 scale as the other criteria, based on the proportion of relevant elementary flows captured in the dataset.
Exchange Category
Each inventory item also needs an exchange category assigned. This determines its base uncertainty variance, following ecoinvent rules.
- Open the DQR panel for an item.
- Scroll to the Exchange Category section below the foreground and background criteria.
- Select the exchange type from the dropdown (e.g. energy, transport, material/product). For pollutant exchanges, also select the emission context: combustion, process, or agricultural.
- XYCLE calculates and displays the base variance (σb²) for that combination.
Repeat for each eligible inventory item.
N.b. not all exchange type and emission context combinations are valid. If a combination is unavailable, select the closest applicable type.
Running the Simulation
- Open the relevant product in XYCLE and click the Visualisation tab in the top left corner.
- Select Uncertainty Analysis from the analysis type dropdown.
- The panel shows a progress indicator confirming how many items have complete uncertainty inputs. Once all items are ready, click Run
- XYCLE runs 1,000 Monte Carlo iterations. The histogram appears within a few seconds.
Reading the Result
The histogram shows the distribution of simulated results for the selected impact category. Use the Monte Carlo Settings panel on the left to switch between impact categories; the default is climate change.
Five summary statistics are displayed above the histogram:
| Statistic | Description |
|---|---|
| Mean | Average of the 1,000 simulation results |
| Median | Midpoint of the distribution |
| P2.5 | 2.5th percentile |
| P97.5 | 97.5th percentile |
| 95% CI | Width of the confidence interval (P97.5 − P2.5) |
Reference lines on the histogram mark the mean, median and percentile bounds. The shaded band between P2.5 and P97.5 shows the 95% confidence interval. Hovering over a bar shows the bin range, count and proportion of runs it represents.
The underlying pedigree factor calculations and exchange-level variances are not shown directly in the UI. For the methodology behind the variance mapping, see the source references at the bottom of this article.
Stale Results
If you modify your DQA after running the simulation, the results are marked as stale. Re-run the simulation to bring them up to date.
Exporting Results
You can export the current view in two formats.
Image (PNG)
- Click Export as PNG in the top right of the histogram.
The exported image reflects the impact category currently selected.
Spreadsheet (XLSX)
Uncertainty results are included automatically when you export your project to Excel. A dedicated sheet—named Uncertainty – {product name}—contains:
- Summary statistics (mean, median, 2.5th percentile, 97.5th percentile, 95% CI) for each impact category
- All 1,000 raw simulation values for each impact category
Methodology References
XYCLE's uncertainty implementation follows the ecoinvent pedigree approach. The key source documents are:
- Weidema, B.P. & Wesnæs, M.S. (1996) — "Data quality management for life cycle inventories—an example of using data quality indicators." Journal of Cleaner Production, 4(3–4), 167–174.