LCAs aren’t deterministic. Your results shouldn’t be either.

In real projects, very few inputs are truly fixed.

• Transport distances vary.

• Supplier data changes.

• Yields fluctuate.

• Emission factors come with measurement error.

Yet many LCAs — and even published EPDs — still report single-point results.

For consultants and verifiers, that creates friction:

• “How sensitive is this assumption?”

• “What’s the expected range?”

• “How confident are we in this number?”

• “Can you document the uncertainty treatment?”

Increasingly, you’re not just asked these questions — you’re required to report them.

Many PCRs, ISO-aligned studies, and verification processes expect:

• uncertainty characterization

• sensitivity analysis

• statistical ranges or percentiles

• documented assumptions

Historically, this meant spreadsheets, manual scenario runs, or external tools.

Now it’s built directly into CarbonGraph.

Introducing Monte Carlo Analysis

CarbonGraph now includes native Monte Carlo simulation across any model or flowsheet.

You can:

1. Assign probability distributions to uncertain parameters

2. Run thousands of randomized scenarios

3. Recalculate full impacts for each iteration

4. Report statistical ranges alongside your results

All using the same impact methods you already use (TRACI, EF, CML, etc.).

No exports. No side tools. No duplicated models.

How it works

Define uncertainty where it matters

Assign distributions to parameters such as:

• transport distances

• scrap rates

• yields

• emission factors

• prices or consumption rates

Supported distributions:

• Fixed

• Normal

• Lognormal (μ/σ or Median/GSD)

• Uniform

• Triangular

• Beta

• Gamma

This lets you match the math to the real-world behavior of each variable.

Run thousands of scenarios

CarbonGraph:

• samples new parameter values

• recomputes the full model

• re-runs all impact methods

• repeats for each iteration

You’re evaluating your actual LCA, not an approximation.

Seed control ensures results are reproducible for audits and verification.

Report defensible outputs

Instead of:

GWP = 1.82 kg CO₂e

You can report:

GWP = 1.82 kg CO₂e
P5–P95: 1.63–2.04 kg CO₂e

Outputs include:

• Percentiles

• Box plots

• Violin plots

• Histograms

• Tabular summaries

These can be directly used in:

• EPD documentation

• verification packages

• sensitivity appendices

• internal QA

What this enables for EPD consultants

✔ Meet reporting & verification expectations

Document uncertainty explicitly instead of relying on assumptions.

Provide:

• ranges

• distributions

• reproducible methods

This reduces back-and-forth during verification.

✔ Identify which inputs actually matter

Quickly see which parameters drive variance.

Focus your data collection where it improves accuracy the most.

✔ Avoid over-interpreting small differences

If two products differ by 2% but uncertainty is ±10%, that’s not a meaningful comparison.

Monte Carlo makes that obvious.

✔ Build more credible studies

Presenting ranges and percentiles signals scientific rigor and transparency — especially for:

• ISO 14040/44 studies

• EPDs

• research-grade assessments

Why we built this

Consultants shouldn’t have to leave their LCA platform to do statistically sound analysis.

Monte Carlo turns CarbonGraph from a single-number calculator into a decision and verification tool.

It helps you:

• defend assumptions

• quantify risk

• and submit stronger, more transparent EPDs

Available now

Open any flowsheet → Uncertainty Analysis → define distributions → Run Analysis

If you want help choosing distributions or interpreting outputs, we’re happy to help.