GHG assessment of an infrastructure project: boundaries, emission factors and reporting thresholds

A lender requests the project's GHG assessment. The technical team responds with a spreadsheet of construction site plant, then stops there. During appraisal, the question becomes more precise: does the project exceed the reporting threshold, which boundaries are covered, on what basis are construction-phase emissions estimated. This article treats the GHG assessment as a compliance deliverable, from triggering the PS3 threshold to the construction estimation method, via the choice of emission factors.

GHG assessment and carbon footprint: distinguishing the two exercises

The two terms overlap, but do not respond to the same requirement. The carbon footprint describes the accounting method, scope by scope, according to the GHG Protocol. A project's GHG assessment is the deliverable expected by the lender: a framed, sourced, comparable quantification that feeds appraisal and monitoring.

The scope-based calculation method is already covered in our article on a project's carbon footprint and its scopes 1, 2 and 3. We do not revisit it here. This article is positioned downstream of the method. It answers three practitioner questions: from what point is an assessment required, which boundaries to adopt, how to estimate emissions when construction has not yet begun.

This distinction has a practical consequence. A good calculation poorly framed misses the requirement. An accurate quantification with an incomplete boundary will be sent back during appraisal. The GHG assessment must therefore be conceived first as a dossier, then as a sum.

The PS3 threshold: what 25,000 tonnes triggers

IFC Performance Standard 3 carries the quantification requirement. It sets a clear threshold beyond which the client must measure its emissions. The standard states: "For projects that produce or are expected to produce more than 25,000 tonnes of CO2 equivalent annually, the client will quantify direct emissions from facilities owned or controlled within the physical project boundary as well as indirect emissions associated with the off-site production of energy used by the project" (IFC PS3, paragraph 8).

Three lessons for the practitioner. First, the threshold applies "annually" and not over the project lifetime. It is read on the representative operating year, not on the project's cumulative total. Second, the wording "produce or are expected to produce" makes the requirement prospective. It is not awaited in the first year of operation: it is anticipated from the design phase, based on an estimate.

Finally, the boundary triggered by the threshold is that of direct emissions and indirect emissions linked to purchased energy. In GHG Protocol vocabulary, this corresponds to scope 1 and scope 2. PS3 does not require, in this paragraph, a complete scope 3. Be careful, however: quantification must cover "all significant sources", including non-energy sources such as methane and nitrous oxide, and land-use changes when they are significant.

The standard also specifies the frequency. "GHG emissions quantification will be conducted annually by the client, in accordance with internationally recognised methodologies" (IFC PS3, paragraph 8). The GHG assessment is therefore not a one-off appraisal exercise. It is recurring monitoring, to be integrated into the project's annual reporting.

Which boundaries to adopt for the assessment

The PS3 threshold sets a regulatory floor. Expected practice often goes beyond, especially for a material-intensive project. The right question is not "what am I obliged to count", but "which boundary makes my assessment legible and defensible".

Three choices structure the framing. The first is the project boundary: what is owned or controlled, hence in scope 1, and what falls under purchased energy, hence in scope 2. This allocation follows GHG Protocol logic and must be explicit from the methodology note.

The second choice is the place of scope 3. For infrastructure, the manufacture of cement and steel, upstream transport of materials and end-of-life weigh heavily. Even if PS3 does not require it under the threshold, an assessment that ignores these items minimises the real footprint. Derived climate frameworks, including the TCFD framework, push to document the value chain. Scope 3 is then treated as an analytical boundary, distinct from the compliance boundary.

The third choice is temporality. An infrastructure project concentrates construction emissions over a few years, then produces operating emissions, or induced traffic, over several decades. Presenting the two phases separately, then cumulatively, avoids mixing a railway profile, heavy during construction and light in use, with a constant-emission profile. This reading joins the logic of a climate analysis integrated into the impact assessment, which repositions the assessment within the alternatives dialogue.

Estimating emissions during the construction phase

The difficulty specific to project assessment lies in its chronology. At the appraisal stage, construction has not begun. There are no fuel invoices, no meter readings, no delivery slips. The construction assessment is therefore an estimate, not a measurement.

The estimate starts from the project's physical quantities, available from the preliminary design stage. Bills of quantities give the tonnes of cement, steel, aggregates. The programme gives the duration and intensity of works. The plant schedule gives the fleet mobilised and its operating hours. These activity data, cross-referenced with emission factors, produce an initial construction footprint.

For direct emissions, two approaches coexist. The consumption-based approach starts from litres of fuel estimated from plant power ratings and operating hours. The ratio-based approach starts from the quantity of work, for example emissions per cubic metre of concrete placed. The former is more accurate when the fleet is known. The latter serves in the upstream phase, when only major quantities exist.

For materials, i.e. the bulk of construction scope 3, the calculation is straightforward: quantity multiplied by material emission factor. Accuracy then depends entirely on the quality of the factors adopted. This is the point on which the assessment's credibility hinges.

An estimate remains an estimate. The dossier gains by displaying its assumptions and a range, rather than a single falsely precise figure. A construction assessment stated "to within plus or minus twenty per cent" is more honest, and more robust during appraisal, than a value to the kilogramme built on generic ratios.

Emission factors: choose, document, trace

The emission factor is the weak link in most assessments. Two dossiers with the same quantities can display very different footprints depending on the factors employed. Lender review therefore dwells less on the addition than on the origin of the factors.

Three families of sources can be mobilised. Generic public databases, such as national carbon bases or recognised material databases, offer default values, broad but traceable. Product environmental declarations from suppliers, when they exist, give specific values, more refined for a given cement or steel. Grid electricity factors, published by international agencies, serve scope 2 and must be chosen for the relevant country and year.

The methodological rule is simple to state. Each factor carries its source, date, geographical boundary. An undated or unsourced factor is an invalid factor under review. The calculation spreadsheet must allow tracing back, for each line, to the exact reference. This traceability is worth more than an additional decimal place.

A final reflex avoids many revisions: geographical and temporal consistency. A European electricity mix factor has no reason to apply to a heavily thermal West African grid. A material factor dated from a decade ago does not reflect current processes. The right factor is the one that corresponds to the project's location, year and process.

What DFIs verify

Beyond the final figure, lenders' climate teams examine the framing and traceability of the assessment.

A GHG assessment is not a calculation, it is a dossier. Three reflexes make it robust during appraisal. Frame the boundary before counting, distinguishing what the PS3 threshold requires from what good practice expects. Assume estimation during the construction phase, with supporting assumptions and range, rather than a falsely precise figure. Trace each emission factor to its dated source.

The right question is not "how much does the project emit", but "which boundary, which data and which factors make this figure defensible". An assessment that answers this question serves climate dialogue and annual monitoring. An assessment that delivers a total without framing goes back to appraisal.