Climate risk, emissions and green bonds, with every number traceable to the data it came from.

EcoCyclus calculates climate risk starting from the location of each plant or facility, builds the emissions inventory and the sustainability reports, and documents the impact of green bonds, all from the same data. For industrial and service organizations.

Every number shows where it came from: the measured data and how it was calculated. In risk results, the platform also states whether the model has been validated in that country, and says so when it has not.

Climate risk

Two plants twenty kilometers apart, in different river basins, can face very different climate risks.

That is why EcoCyclus starts from the location of each asset, not from the country average. For each hazard it uses the closest data to the asset that exists (the point, the basin, a grid cell, the weather station or the nearest reference city); when only country data is available, it uses that and says so. In Latin America, where available data tends to be very coarse, that difference changes the conclusion.

Much as an insurer would, the platform estimates for each asset how much can be lost per year on average from each hazard (average annual loss). From that it assigns each hazard a rating from A to G.

For river and coastal flooding, the hazard is measured at the asset's point. For earthquake and wind, at the nearest reference city. For all four, damage is calculated with each asset's value, type and floor area.

For floods it also estimates how much would be lost in an extreme event, one with a 1 in 250 chance of occurring in any given year (probable maximum loss). When the available models disagree, the result shows that range instead of hiding it.

Loss in money and rating

River flooding, coastal flooding, earthquake and wind.

Exposure and trend, not converted to money

Extreme rainfall, drought, water stress (how much of the available water is already being used) and wildfire danger.

For extreme rainfall, there is still no reliable way to go from rainfall intensity to damage. Water stress, drought and wildfire, on the other hand, measure how exposed a place is, not how much damage it causes, and so they are not converted to money.

For both types of flooding, risk can also be compared with and without a protective structure, under the current climate. This is not available for earthquake and wind.

It also assesses transition risk (the impact of new regulations, carbon prices or market shifts) using the central banks' reference scenarios (NGFS). It is shown separately from physical risk, because adding them into a single number would produce a misleading figure.

On the same basis it runs a first-pass climate risk assessment of suppliers (an approximation, not an insurance calculation).

For projects pursuing LEED v5 certification, it produces the climate risk assessment report in the format the certifier requires. It covers part of the hazards LEED requires; the details are in How it works.

What it does not do

  • The rating is per hazard, with no combined letter per asset, and it requires declaring the asset's value; currency conversion covers 30 currencies, but not yet the Argentine peso.
  • Earthquake and wind are calculated only for assets with a reference city within 100 km and whose asset type has a damage curve; otherwise, there is no loss calculation for that hazard. Wind, in addition, is calculated only in the US, and outside the US earthquake loss is indicative (screening level).
  • Precision varies by hazard: nearby assets may share the hazard of the same reference city, and for several hazards the only data available is at country level.
  • It covers a declared set of countries, not the whole world.

How risk is calculated, in “How it works”

Why trust it

Many platforms promise that their numbers can be traced back to their source. In EcoCyclus that is built into the way data is stored.

It works like a bank statement: every balance is explained by its transactions, and an error is not erased, the correction is recorded. From any indicator you can reach the calculation, the factor used and the measured data. That lets an auditor, an investor or an external reviewer verify it.

Illustrative example
  1. Total plant emissionst CO₂e for the period
  2. Diesel consumption behind itliters measured in the period
  3. Emission factor usedvalue, source and version

In risk results, the platform states whether the model that produced them has been validated in that country; when it has not, it says so. A system that cannot show what is missing cannot prove what it has either.

  • Correcting a data point does not delete the previous one. Both remain, and you can see which replaced which and when.
  • If a formula changes, earlier results are still explained by the formula used at the time.
  • A model validated in one country is not thereby validated in another, and the platform states this country by country.
  • Each result states the method used to obtain it and the quality of its source data. Each organization's data is kept separate from everyone else's.

What it does not do

  • The platform itself is not audited or certified. What it shows is the method and the data behind each figure and, where applicable, who is accountable for it.

How traceability is built, in “How it works”

Emissions

Your organization's emissions inventory, under the two most widely used standards, with every number explained.

It calculates three groups of emissions: direct ones, such as what comes out of your stacks and vehicles (Scope 1); those from the electricity and energy you buy (Scope 2); and those from your value chain, such as what you buy from your suppliers (Scope 3; the inventory supports all 15 GHG Protocol categories). The same inventory is presented under the GHG Protocol or under ISO 14064-1, without redoing calculations.

Electricity emissions are calculated in two ways: based on the local power grid and based on the company's energy contract. Both are kept: the GHG Protocol requires reporting both, and other frameworks ask for one or the other.

Electricity consumption kWh measured Based on the power grid local grid factor Based on the energy contract contracted supplier's factor
Both figures are kept: the GHG Protocol requires reporting both; other frameworks ask for one or the other.

From any number you can reach the calculation, the emission factor used (how much is emitted, for example, per liter of diesel or per kWh of electricity) and the measured data it came from.

The calculations do not use artificial intelligence: they work like a calculator. With the same data, the result is always the same and can be verified.

Standards, factors and inventory method, in “How it works”

Product footprint

The carbon footprint of your products, shareable with your customers without exposing your internal data.

It calculates a product's emissions from raw material extraction until it leaves the plant (“cradle to gate”), under ISO 14067. The footprint can be calculated from the same data as the company's inventory, and in that case the two numbers are consistent by design.

A supplier can publish its product's footprint and its customer can use it directly in its own calculation. If the supplier updates it, the customer sees that a new version is available.

Supplier Consumption, recipes, internal factors Published footprint internal data does not cross over Customer Uses the footprint in its own calculation and sees if there is a new version

It works like the nutrition label on a package: the customer sees the result, not the recipe. The customer never accesses the supplier's internal data, and the system guarantees this, not a usage rule.

For projects pursuing LEED v5 certification, it also calculates the emissions associated with the building's construction materials.

What it does not do

  • It measures only until the product leaves the plant (not its use or disposal), and only its climate impact (greenhouse gases), not other environmental impacts such as water use.

How the footprint network works, in “How it works”

Reporting

One sustainability report, checked against several standards at once.

Many companies must report under several standards: the European ones (ESRS), the most widely used global standard (GRI) and the international climate disclosure standard for investors (IFRS S2). The platform shows how much of each standard your report already covers and what is missing. Where ESRS and GRI ask for the same thing, the information is entered once and serves both.

The texts written for the report (policies, explanations) are stored with their change history and reused across standards.

It links indicators to the Sustainable Development Goals using GRI's official table, and assesses which activities could fall under the European Union's classification of sustainable activities (EU Taxonomy).

ESRS · GRI · IFRS S2 · EU Taxonomy Art. 8 · ISO 14064-1 · ISO 14067 · GRI 303 · ICMA GBP · EuGBS

What it does not do

  • It indicates which activities are eligible under the EU Taxonomy, but does not claim they meet all its requirements.
  • It does not calculate an SDG “compliance percentage”, because those goals are measured by country, not by company.

How coverage is assessed, in “How it works”

Green bonds

Green bonds for adaptation: measuring what no one has yet defined how to measure.

Whoever starts measuring now arrives with a documented method before the rule is written.

Green bonds finance two kinds of projects. Mitigation projects reduce emissions, and established indicators already exist to measure them. Adaptation projects protect assets from the effects of climate (for example, flood defenses), and for them ICMA's impact reporting guidance does not yet set core indicators: it only gives examples.

If your company has or is going to issue a green bond, EcoCyclus records how each amount was used, links each project to the asset it protects and generates the reporting document for the investor or the external reviewer.

Bond Use of proceeds Asset Project Result

To measure impact, it compares the asset's risk with the project against the risk it would have without it, like comparing the same plant with and without a levee. That calculation comes from the risk model, based on the design level the company declares for the project, not from a manually entered number. Today this comparison is available for river and coastal flooding, under the current climate.

It works with the two most widely used green bond standards: ICMA's Green Bond Principles and the European Green Bond Standard (EuGBS). A single project can be declared as protection against several hazards.

What it does not do

  • The comparison assumes the project withstands every event up to the level it was designed for: it does not account for the defense failing or the water overtopping it. That is why the benefit it shows is the best case, and only for floods under the current climate.
  • For the EU Taxonomy it indicates eligibility, not full compliance.
  • EcoCyclus prepares the document; the bond's certification is issued by an external reviewer.

Method of the with-and-without comparison, in “How it works”

Risk management

Your organization's risk management, organized according to ISO 31000.

A risk register with categories, assessment criteria and a risk matrix. Each risk is assessed in its own entry, and the register is reviewed in periodic cycles.

Risk register and administration, in “How it works”

Want to see how it would work with your data? Write to us.

info@ecocyclus.app

EcoCyclus is built by Fernando Amar, from the Province of Buenos Aires, Argentina. The method comes from practice in environmental and climate management, and from it comes the platform's rule: when the data is not enough to support a figure, the gap is not filled with an assumption.

And it is built from Latin America, where national averages hide more differences between one place and another.

Home / How it works

How it works

This page is for those who want the detail: reviewers, auditors, risk or sustainability analysts. It explains how each result is calculated, what data it comes from and where the limits are. Technical terms are defined the first time they appear; full definitions are in the glossary.

Climate risk

The risk engine, hazard by hazard.

The method

Underneath is a probabilistic engine. For each hazard it builds an exceedance curve: the relationship between the size of a loss and the annual probability that it is equaled or exceeded. Two indicators come from that curve:

  • Average annual loss (AAL). What would be lost on average per year, considering all possible events and their probability.
  • Probable maximum loss (PML-250). The loss in an event with a 250-year return period, that is, with a 0.4% probability of occurring in any given year. It is published as an indicator for river and coastal flooding. For earthquake and wind it is calculated within the curve, but not published as an indicator.

When there are several models for the same hazard, the uncertainty between them is propagated to the result instead of being averaged away early. Coastal loss is calculated with a declared uncertainty band, under RCP emissions scenarios.

Damage is calculated with each asset's own data: its value, its type and its floor area. For earthquake, the platform uses building fragility curves (the probability of damage given the intensity of ground shaking), from the HAZUS methodology.

Where the hazard is read

The hazard is taken from the closest data to the asset that exists for each hazard. That distance varies by hazard, and the table shows the maximum allowed:

HazardWhere the hazard is read
River and coastal floodingAt the asset's point, at a maximum distance of 1.5 km.
Earthquake and windAt the nearest reference city, up to 100 km away. An asset with no reference city within that radius has no loss calculation for that hazard, and loss is calculated only for asset types with a damage curve (for example, there is no seismic curve for oil and gas extraction facilities). In Argentina there is a single seismic reference point (Neuquén): assets within 100 km share its curve, and those farther away have no earthquake loss calculation.
Water stress and droughtIn the river basin that contains the asset's point.
Wildfire dangerIn the nearest grid cell, up to 35 km away.
Extreme rainfallAt the nearest weather station with a long record, up to 50 km away; the distance is reported. Available where there is a station within 50 km.
Transition riskBy country.

For other hazards in the location panel, the only data available is at country level, and the platform says so.

The A to G rating

The rating is the average annual loss divided by the asset's declared value. It is assigned per asset and per hazard; there is no combined letter across hazards. It applies to the four hazards with loss in money: river flooding, coastal flooding, earthquake and wind (the latter only in the US).

LetterAverage annual loss / asset valueIn other words
Aless than 0.10%Loses on average less than 1 in 1,000 of its value per year.
B0.10% to 0.25%Up to 1 in 400 per year.
C0.25% to 0.50%Up to 1 in 200 per year.
D0.50% to 1%Up to 1 in 100 per year.
E1% to 2.5%Up to 1 in 40 per year.
F2.5% to 5%Up to 1 in 20 per year.
G5% or more1 in 20 or more per year.

With no declared value there is no letter. If the loss and the value are in different currencies, the platform converts them pairwise and attaches the rate, source and date to the result. Conversion covers 30 currencies; the Argentine peso is not yet included. If the currency cannot be converted, there is no letter, and the platform says why.

Hazards not converted to money

  • Extreme rainfall. A damage function is still missing: a reliable way to go from rainfall intensity to damage to the asset.
  • Drought. It is a composite index that already incorporates the vulnerability of the place. Converting it into loss would count that vulnerability twice, so it is not done.
  • Water stress. It is the ratio between the water used and the water available, not a physical intensity. There is no damage function that converts it into loss.
  • Wildfire danger. It is a meteorological danger index, not a measure of damage.

These hazards are reported as exposure and trend.

The with-and-without comparison

Hazards are classified into three levels according to what they support:

  • Support the with-and-without comparison: river and coastal flooding.
  • Calculated in money but do not support the comparison: earthquake and wind.
  • Can only be declared: extreme rainfall and erosion.

The details of the method and its limits are in green bonds.

Transition risk

Transition risk is assessed with the scenarios of the Network of Central Banks and Supervisors for Greening the Financial System (NGFS). Physical and transition risk are not added into a single number: they do not share a unit of measure, and collapsing them would produce a figure that neither supports.

Suppliers

On the same basis runs a climate risk assessment of suppliers. It is screening level, not an actuarial calculation.

LEED v5: resilience assessment (IPp1)

The same engine produces, asset by asset, the climate risk assessment report in the format of USGBC's official template for the LEED v5 IPp1 prerequisite.

It covers part of the twelve hazards IPp1 requires.

Sources

WRI Aqueduct · HAZUS · USGS · STORM · NOAA GHCN-Daily · INFORM Risk · JRC · NGFS · ThinkHazard! · World Bank Carbon Pricing · GEFF/CEMS

Each source is recorded in the figure it feeds. Sources are cited by name, without logos: using a public dataset is not a relationship with its publisher.

What it does not do

  • Outside the US, earthquake loss is indicative, screening level.
  • Wind (hurricane) is calculated only in the US.
  • The with-and-without comparison runs under the current climate, not under future scenarios.
  • The geographic domain is a declared set of countries, not global coverage.

Why trust it

Auditability is not declared in a document: it comes from how the data is stored.

A platform can promise traceability and be unable to support it in its database. In EcoCyclus traceability is the consequence of three construction decisions: how the measurement is stored, how formulas are versioned and how the validation level is resolved at the moment someone reads the figure.

Data Factor Calculation Figure Validation level enteredversionedversionedresulting
The validation level enters from the side: it does not change when the figure is recalculated; it changes when a new calibration of the model is published for that source and that geography.

Three different things a figure can declare

  • Model validation. Applies to risk results. It is per source-and-geography pair: a model checked in one country is not validated in another. Today most of the portfolio is shown as not validated, and the platform shows it instead of assuming otherwise.
  • Source data quality. Each data point carries its quality classification.
  • Assurance (N1 to N4). Who is accountable for the figure. It is declared by the issuer, not by the platform, and exists in project results (see green bonds).

Properties

  • Correcting a data point does not delete the previous one: both remain, and you can see which replaced which and when.
  • Each calculation is tied to the formula version it was run with. If the formula changes, the earlier result is still explained by the one that produced it.
  • From an indicator you reach the calculation, from the calculation the factor, and from the factor the measurement that fed it.
  • Each result carries the method used to obtain it. Each organization's data is kept separate from everyone else's.

What it does not do

  • EcoCyclus is not audited or certified as a platform. The assurance level is something the issuer's data declares, not a status of the platform.

Emissions

The inventory, its standards and its factor base.

Two standards, one inventory

The inventory is issued under the GHG Protocol or under ISO 14064-1. It supports all 15 GHG Protocol Scope 3 categories and regroups them into ISO 14064-1 categories 3 to 6 without recalculating.

Scope 2 by both methods

Electricity emissions are calculated by location (the factor of the grid where the asset is) and by market (the factor of the purchase contract). Both figures are kept: the GHG Protocol Scope 2 Guidance requires reporting both where market instruments exist, and other frameworks ask for one or the other.

Factor base

The factor base can be searched by several fields at once and records which source each factor came from. It converts gases to CO₂ equivalent with the declared global warming potential (GWP) version, normalizes units and resolves the factor that applies to each geography.

For Scope 3, the library includes reference factors from DEFRA (EEIO) and the EPA (spend-based, by NAICS) for the upstream categories and some downstream ones. For the rest, the calculation uses supplier data or a declared factor.

Thresholds and calculation

Regulatory thresholds are evaluated on the inventory itself, not on a copy. The calculation engine is deterministic: same data, same result. The platform uses artificial intelligence to assist in drafting narrative texts, not in the calculations.

Product footprint

Product footprint and the footprint network between organizations.

The product footprint is calculated cradle to gate (until the product leaves the plant) under ISO 14067, with the study report and the allocation method aligned with the standard. The separation between fossil and biogenic carbon is enforced by the data model, not by a usage convention.

A study can draw on declared inputs, on footprints published by suppliers or on catalog factors (ÖKOBAUDAT, DEFRA). When it runs on the company's inventory inputs, reconciliation with the inventory holds by construction, not through a later adjustment.

The network between organizations

One organization publishes the resolved footprint of its product and another uses it in its calculation; if more than one footprint is available for the same input, precedence rules define which is used. When the supplier publishes a new version, the platform detects it by the identity of the publication and whoever was using it sees that a new version is available. Only the published footprint crosses over: separation between organizations is enforced by the system, not by convention.

LEED v5: materials carbon (MRp2)

The same engine calculates the embodied carbon of a project's construction materials, by location, with a catalog under EN 15804 (ÖKOBAUDAT) and DEFRA.

It works with industry-average values, not with manufacturer-specific verified environmental product declarations (EPDs); that portion of the score is covered together with a specialized partner.

What it does not do

  • Cradle to gate only and climate change only: it does not cover the product's use or end of life, or other environmental impacts.

Reporting

Coverage across several frameworks at once, without imposing a hierarchy among them.

The coverage engine assesses the report against several frameworks at the same time. The equivalences between ESRS and GRI live inside the data model, not in a separate spreadsheet. Qualitative responses (policies, explanations) are an entity of their own: they are versioned and reused across frameworks.

Linking to the Sustainable Development Goals uses the official GRI–SDG correspondence table. The indicator catalog is navigated by ESRS topic. EU Taxonomy eligibility is assessed from the economic activity code (NACE).

ESRS · GRI · IFRS S2 · EU Taxonomy Art. 8 · ISO 14064-1 · ISO 14067 · GRI 303 · ICMA GBP · EuGBS

What it does not do

  • EU Taxonomy: eligibility only. It does not claim alignment or assess the “do no significant harm” (DNSH) criterion.
  • SDGs: it does not issue a compliance percentage. This is deliberate: SDG indicators are defined by country, not by company.

Green bonds

From bond to result: allocation, measure and comparison.

The reference framework

ICMA's impact reporting guidance (Handbook – Harmonised Framework for Impact Reporting) offers only example indicators for adaptation, with no core indicators. That remains the case in the current edition, from June 2026. The platform works with ICMA's Green Bond Principles (GBP) and the European Green Bond Standard (EuGBS) at the same time.

Versioned regimes

Eligibility is declared under a named, versioned regime, not as a fixed category. Two regimes can govern the same result at the same time, and the previous version remains available.

Allocation and measure

The proceeds allocation register keeps its full history: reallocating an amount does not overwrite the previous allocation. Each asset is linked to the declared adaptation measure, which states whether the project is planned, under construction or completed. A single project can be declared against several hazards, but its benefit is calculated only for the two flood types.

When and who

They are declared separately. When: whether the result is an estimate before the project (ex ante) or a measurement after it (ex post). Who is accountable, on the assurance scale:

LevelWho is accountable for the figure
N1Self-declared: reported by the issuer without review.
N2Internal verification: reviewed by a function within the issuer.
N3External verification: reviewed by an independent third party.
N4Certification: issued by an accredited verifier, for example the EuGBS external reviewer.

The level is declared by the issuer; the platform does not grant it.

The with-and-without comparison

The comparison (counterfactual) is ex ante, against the “no action” scenario. The risk with the project is derived from the model based on the design level the company declares for the project; it is not entered manually. The model treats events up to the design return period as causing no loss, because it assumes the defense fully contains them.

What it does not do

  • It does not model breach or overtopping of the defense, or its fragility (the probability that it fails given the load of the event). That is why it overestimates the benefit: the result is a maximum, not a central estimate. This is resolved with a fragility curve from a published source.
  • It runs under the current climate scenario, with the mean value; not under future scenarios.
  • EU Taxonomy: eligibility only, not alignment or DNSH.
  • EcoCyclus does not certify green bonds: it produces the document, and certification is issued by an external reviewer.

Risk management

Risk register and administration.

The risk module follows ISO 31000 nomenclature: categories, assessment criteria and review cycles are entities in the model, with a risk matrix and a statement per entry.

Each organization has its own administration console and manages its factor base on a self-service basis.