Dawn
Tutorials/EPD Generator review

Reviewer checklist (four-eyes review)

The checklist that the reviewer completes for each EPD from the EPD Generator before it goes to EPD Hub: per module, the bill of materials, quantities, transport, energy, reference year, free variables, plausibility against the main drivers, and the issues to clear.

Reviewer checklist (four-eyes review)

This reviewer guidance is for the reviewer. The reviewer is the second member of the organisation, who examines an EPD before it goes to EPD Hub. The reviewer has the EPD reviewer role, and did not prepare, edit or submit the EPD. The reviewer is not a verifier. The review does not replace the third-party verification by EPD Hub.

The checklist follows the model of the EPD Danmark checklist for EPD tools. For each module, it examines the bill of materials (BoM), the quantities and units, and the transport. It also examines the energy, the reference year of the data and the free variables. It adds the checks that the tool itself does not make.

NOTE: ECO Platform does not ask for a four-eyes review when the EPD shows all variable input data (ECO TVG 4.3). Dawn asks for it because some inputs, for example the waste data sets, are not printed in the EPD.

Before you start

  1. Get the product name and the EPD draft from the user. The EPD shows "In review by" and your name.
  2. Get the evidence of the user: BoM, drawings, utility bills, waste statements, sales statistics.
  3. Open the product in the tool. Keep the free-variable reference open. It gives the allowed range and the evidence of each input (P-ID).
  4. Make a copy of the review record (at the end of this page).

Do each check. Mark it "OK", "Not OK" or "Not applicable". For each "Not OK", write a finding: the check ID, what is wrong, and where.

G. General

IDCheckWhere in the toolOK when
G1The tool version is verified and its "Valid until" date is not past.Settings › EPD toolThe tool version shows its verification.
G2The product name, description and declared unit agree with the data sheet.EPD tab › ProductThe values agree. The declared unit is the one of the PCR, or the one that your organisation selected. For 1 m², the width and height on the Info tab › Dimensions card are those of the modelled product.
G3The mass per declared unit agrees with the weighed or drawn mass of the real product.EPD tab › ProductThe difference is small and explained.
G4The declared modules are correct. A module declared as 0 has a reason.EPD tab › CalculationA1–A3, C1–C4 and module D are declared. Each module "declared as 0" is true for the product.
G5The data period is 12 months and is correct.EPD tab › EPD detailsAll primary data of A3 are from this period.
G6Each text agrees with the model and the results.EPD tab › DescriptionsNo text states a number, a material or a process that is not in the model. No warning "Descriptions: generated from other facts". If it shows, the facts changed after the texts were generated. Generate the texts again.
G7The method fields are complete: PCR, programme operator, background database with its version, LCA software.EPD tab › MethodEach field has a value. The database field names the ecoinvent version.

A1–A3. Product stage

IDCheckWhere in the toolOK when
A1-1BoM complete. Each component of the BoM or drawing is in the model.BOM tabNo component is missing. Small parts left out are inside the cut-off rules and named.
A1-2Quantities and units. Each quantity is in the unit of the part. Pieces, kg and m are not mixed up.BOM tabEach quantity agrees with the BoM.
A1-3Own parts. Each own part has a mass, and its composition totals 100 %.Part › material tableNo mass is empty. Each total is 100 %.
A1-4Conversions. Masses converted from area, length or pieces are correct.Part, BoM evidenceThe conversion is shown in the evidence.
A1-5Materials. Each material is the nearest library material, with the correct recycled content.Part › material tableRecycled and virgin library materials are combined to give the real recycled content.
A1-6Production loss. Each loss % agrees with the scrap or yield records.Production tab › Production lossEach value has evidence.
A2-1Transport mode and distance. Each bought part has a route from each supplier. The mode and the distance agree with the supplier country.Part › Supply chainNo bought part is without a route. Each distance is plausible.
A2-2Supplier shares total 100 % for each part with several suppliers.Part › Supply chainThe total is 100 %. The tool does not check this.
A3-1Energy complete. Electricity, heat, fuels, compressed air and water of the site are all entered.Production tab › Utilities, Locations › UtilitiesEach energy carrier on the utility bills is in the model.
A3-2Energy data set. Each utility is of the country of the site. Each utility is the correct carrier.SameThe electricity is of the country of the site. A utility does not use the data set of a different carrier.
A3-3Same data year. All site data are from the data period of the EPD.LocationsAll totals are from the same 12 months.
A3-4Allocation. The allocation basis and the annual volumes are correct.Locations › AllocationsThe basis has a reason. The volumes agree with the production statistics.
A3-5Waste streams. Each waste stream has an amount, route shares of 100 %, and data sets of the system model of the tool version.Locations › Production wasteThe tool does not check these data sets when the EPD is submitted for review. Each data set has a reason.
A3-5aProduction scrap not counted twice. The finding "Production scrap may be counted twice" shows when a material has a BOM production loss and a site waste stream with the name of the same material.Impact › Resource use › FindingsNo such finding, or the two entries are different scrap and the reason is written. If they are the same scrap, ask the user to remove the site waste stream or to set the BOM production loss to 0.
A3-6Packaging. Each packaging item and quantity agrees with the packing list.Production tab › PackagingThe quantities agree.

A4–A5. Construction process stage

IDCheckWhere in the toolOK when
A4-1Transport mode and distance to the building site agree with the main markets.Use tab › Transport to customerThe distance agrees with the sales statistics.
A4-2Changed A4 values on the EPD tab have a source.EPD tab › EPD details › A4 cardEach "Edited" value has a source.
A5-1Installation loss has evidence.Use tab › Installation lossThe value agrees with installer data.
A5-2Installation inputs are per declared unit, not per year.Use tab › InstallationEach amount is for one product.
A5-3Only A5 content in A5. A5 contains installation, installation loss and packaging waste only.Impact tab, results per moduleNo product-stage process, for example transport to the factory gate, is booked in A5.

B1–B7. Use stage

IDCheckWhere in the toolOK when
B-1RSL has a basis and, where the tool asks for it, a justification.Use tab › Expected LifetimeThe basis has evidence. With the EN 17213 feature, an RSL over 30 years has the ISO 15686-8 data record.
B-2Exchanged parts are in the correct category. Worn parts are in maintenance (B2).Use tab › Spare PartsA part in replacement (B4) has the deviation justification.
B-3Exchanged parts have an impact. No exchanged part shows "no impact report". A spare part with no data set, no published LCA and no material with a data set books nothing.EPD tab, issuesNo issue "Use stage: spare part without an impact report". For EN 15804 organisations, this issue blocks "Ready for verification".
B-3aMass of exchanged parts. The material mass of one exchanged unit is plausible.EPD tab, issuesNo warning "Use stage: exchanged mass not plausible", or the warning has a reason.
B-4Per year, per exchange and total amounts are entered in the correct basis.Use tabEach amount agrees with the basis of its card.
B-5Operational energy (B6) uses the electricity of the country of use.Use tab › Operational energyThe country agrees with the place of installation and use, or the EPD names the mix.
B-6Each B input is booked. No B input shows "Not mapped: books nothing".Use tab, EPD tab issuesNo "unmapped" or "invalid" issue on a B module.
B-7Deviations. A changed exchange count and a worn part in B4 have a justification. With the EN 17213 feature, each deviation from the default scenario also has one.EPD tab › use-stage menu (shown only with the EN 17213 feature) and deviation justificationThe justification is written and true.

C1–C4 and module D

IDCheckWhere in the toolOK when
C-1End-of-life shares have a source for the market of the product.End of Life tabChanged shares have a source. Default shares fit the market.
C-2C1 is entered, or "declared as 0" has a reason.End of Life tab › DisassemblyThe reason is in the C1–C4 text.
C-3C2 transport mode and distance are plausible.End of Life tab › Transport to wasteA route is entered.
C-4C3 processing data set has a reason and the system model of the tool version.End of Life tabThe tool does not check this data set when the EPD is submitted for review.
D-1Module D has the expected sign, and each Q that is not confirmed is known.Impact › Resource use › Module D net flowModule D is a credit for recycled metals. "Q not confirmed" is accepted, or Dawn has confirmed Q.
D-2Findings of the resource use are examined.Impact › Resource use › FindingsNo finding with severity "error". Each warning has a reason.

F. Free variables described

IDCheckWhere in the toolOK when
F-1Each value that the user changed from a default has a source.EPD tab, badge "Edited"Each "Edited" value has a source in the evidence.
F-2Each justification asked for by the tool is written and true.Use tab, EPD tabNo justification is empty or general.
F-3Each variable input that the EPD does not show is in the evidence.Free-variable referenceWaste data sets (P-18, P-31), supplier shares (P-11) and site volumes (P-15) have evidence.

P. Plausibility against the main drivers

IDCheckWhere in the toolOK when
P-1The main drivers of the results are the expected ones.Impact tabThe largest contributors of GWP-total agree with the main drivers in the list below. Each main driver has evidence.
P-2The results per kg of product are in the range of similar products.Impact tab, a reference EPDA difference of 10 % or more in a core environmental impact indicator has a reason. Look first at the main drivers below.
P-3A1 is the largest part of A1–A3 for a product of bought materials.EPD tab › CalculationA large A3 has a reason, for example energy-intensive processing.
P-4PENRT and ADP-fossil agree within 10 % for each module (EPD Hub PCR 4.5.3).Impact › Resource useNo finding V-PL.

Main drivers

The tool project report, section 7, gives the sensitivity assessment of the tool on a door and a window. A parameter is a main driver when one of two rules applies:

  1. A change of 10 % moves GWP-fossil A1–A3 or GWP-total A1–C4 by 1 % or more, or another core environmental impact indicator by 5 % or more.
  2. A change in the tested range moves a core environmental impact indicator by 10 % or more (EN 15804+A2, 9, NOTE, and EPD Hub GPI 3.7).

Examine these inputs first. A small error in them gives a large error in the EPD.

InputWhy it is a main driverWhat to examine
BoM quantities (P-03) and masses of own parts (P-07)A1–A3 changes almost in proportion to the mass. On the pilots, all BoM lines +10 % gave GWP-fossil A1–A3 +9.5 %. Module D follows the mass of aluminium.The quantities and masses of the heaviest parts agree with the BoM and the drawings.
Material of a heavy part (P-08)A change from aluminium to PVC or steel moved GWP-fossil A1–A3 by −53 % to +58 % and module D by −244 % to +76 %.Each heavy part has the correct material and recycled content.
Material shares in a part with two materials (P-08)The aluminium spacer share of an IGU moved GWP-fossil A1–A3 by up to +15 %.The shares agree with the supplier data sheet.
RSL (P-25)The use stage gives about two thirds of GWP-total A1–C4. RSL +10 % gave GWP-total A1–C4 about +5 %.The RSL has evidence. With the EN 17213 feature, an RSL over 30 years has the ISO 15686-8 data record.
Operational energy use in B6 (P-26) and its electricityB6 energy +10 % gave GWP-total A1–C4 about +5 %. The electricity mix of B6 moved GWP-total A1–C4 by −52 % to −4 %.The amount per year has a source. The electricity is of the country of use.
Exchange of worn parts (P-27)One more exchange of the IGU gave GWP-total A1–C4 +17 % and module D −34 %.The number of exchanges, the service life of each component and the quantity per exchange have evidence.
End-of-life route shares (P-28)The shares moved module D by −29 % to +81 %.Changed shares have a source for the market of the product.
Installation loss (P-23)25 % installation loss gave GWP-total A1–C4 about +19 %.The loss agrees with installer data.
Production loss (P-04)A loss near 100 % doubles A1–A3.Each loss agrees with the scrap or yield records.
Inbound transport mode (P-12)Air freight of one part gave GWP-fossil A1–A3 +4.9 % and EP-terrestrial A1–A3 +13 %.Each mode agrees with the supplier route. Air freight has a reason.

NOTE: The value-correction factor Q and the point of substitution also move module D by more than 10 %. Dawn sets them on the library materials, and the user cannot change them there. The user can enter Q on the benefits of its own site waste streams (P-52, P-18). On a stream in the end-of-life phase, Q changes module D. Check that each such Q has a justification.

I. Issues to clear before "Submit for review"

"Submit for review" refuses an EPD with an issue of the kind "missing" or "invalid", except the file uploads. The tool does not stop on warnings or on findings of the resource use. The reviewer does.

IDIssueWhat to do
I-1An issue of the kind "missing" on a necessary fieldAsk the user to enter the value.
I-2An issue of the kind "invalid"Ask the user to correct the input.
I-3An issue of the kind "warning"Accept it with a written reason, or ask for a correction.
I-4A declared module whose inputs are not bookedAsk the user to remove the input or to undeclare the module. Contact Dawn if a module mapping is missing.
I-5A finding with severity "error" in the resource useAsk the user to correct the input.

Review result

  • Accepted. All checks are "OK" or "Not applicable", and all warnings have a reason. Select Approve. The tool seals the EPD record. Then a member can send the EPD to verification.
  • Not accepted. Select Return for changes, and write the findings as the reason. The draft unlocks. The user corrects the input and submits the EPD again, and the review starts again.

Review record

The tool does not store the checklist yet. It keeps the review history of the EPD: who submitted, approved or returned it, when, and the reason. Keep the completed record with the evidence of the EPD for the period of validity of the EPD.

FieldValue
Product and variant
EPD draft or EPD number
Tool version
Prepared by (user)
Reviewed by (reviewer)
Date of the review
Time used
Checks "Not OK" (IDs)
Findings and how they were cleared
Warnings accepted, with the reason
ResultAccepted or Not accepted
Signature of the reviewer

Subscribe to our newsletter